AI monitoring and analytics system for Prozorro tenders + Telegram Bot

Budget: 27000 UAH Deadline: 30 days

Good day,

I read the specifications carefully. This is not a Telegram bot — it is an ETL platform with a bot as one of the outputs. That’s how I will build it.

— — —

🔧 STACK

— — —

Python 3.12 + FastAPI (async)

PostgreSQL + SQLAlchemy 2.0 + Alembic

aigram 3.x + FSM for bot commands

Celery + Redis — queue for polling Prozorro

Docker Compose — the entire project in one command

pgvector — directly in the schema, for future AI search

— — —

🏗 MVP ARCHITECTURE

— — —

Fetcher-worker → incremental synchronization by dateModified (not a full scan — this is critical for speed)

Deduplication → composite PK: tender_id + revision + lot. Change history does not "flow"

Filtering → at the PostgreSQL level: GIN indexes on CPV array, Full-Text Search on keywords. Not in Python, but in the database — fast and scalable

Two workflows at once:

• Active tenders (submit an application)

• Contracts/leads (who has already purchased)

Analytics → a separate layer of materialized views: winners, % price reduction, statistics on competitors' EDRPOU, customer statistics. Does not overload the main backend

Telegram bot → lightweight client over API: /start /settings /keywords /cpv /region /last — through FSM aiogram 3

Deduplication of notifications → notifications table: one push for a pair of user_id + tender_revision. No spam

— — —

📊 WHAT YOU WILL GET IN MVP

— — —

✅ Integration of Prozorro API (tenders + contracts)

✅ Filtering: CPV, keywords, region, amount, customer, deadline

✅ PostgreSQL with change history and protection against duplicates

✅ Telegram bot with all commands and FSM settings

✅ Analytics: winners, participants, % reduction, competitors by EDRPOU

✅ Docker Compose — deployment with one command

✅ pgvector in the schema — ready for AI without rewriting

✅ Alembic migrations + pytest coverage of critical modules

— — —

🚀 AFTER MVP (already embedded in the architecture)

— — —

AI analysis of tender documentation through OpenAI + RAG

Affiliation search through EDRPOU connections

Web dashboard

CRM logic and task setting

— — —

💰 COST AND TERMS

— — —

MVP: 32,000 UAH

Term: 3-4 weeks

Why not 25,000 like everyone else? Because the right architecture from the start costs more than rewriting twice. I provide an MVP with a solid foundation for a reasonable price.

— — —

❓ TWO QUESTIONS FOR YOU

— — —

1. Are CPV codes fixed or does the user add them through the bot? This affects the data model.

2. Is historical loading of tenders for 1-3 years needed for analytics, or do we start from the current moment?

Ready to start after your response. Expand

Budget: 27000 UAH Deadline: 20 days

Good day.
I am implementing a full-fledged MVP system on Node.js + PostgreSQL: integration with the Prozorro API with filtering by CPV codes, region, keywords, and amount, a Telegram bot with all necessary commands, as well as an analytical module on winners and competitors - I have practical experience working specifically with the Prozorro API, having processed about 30 million records within a similar project. I will deploy the system in Docker with the correct architecture, which will allow for seamless scaling to AI analysis and a dashboard in the next stages. I have also integrated the OpenAI API not only for content generation but also for analyzing the obtained data.

I indicate the estimated cost and terms for the MVP in the bid.

I suggest discussing the price and terms in private messages.
I am ready to discuss the details and start working. I would be happy to collaborate.

Budget: 25000 UAH Deadline: 25 days

Hello, my name is Mykyta.
I am ready to take on the development of your AI monitoring system for Prozorro with a Telegram bot with great pleasure and responsibility.

I have 2 years of experience in developing complex backend systems and integrating with APIs, particularly in data parsing and monitoring.
I work with Python, FastAPI, PostgreSQL, Docker, which are the key technologies for this project, and I have dealt with analytical systems.
I understand the principles of working with the Prozorro API and the possibility of integrating AI functionality at later stages.
Portfolio: Freelancehunt

Estimated price, timeline for MVP: 25,000, 25-30 days.
I am ready to discuss the details and propose an optimal architectural solution.

Budget: 25000 UAH Deadline: 7 days

Hello!

I develop bots for Telegram using NodeJS. I have experience with the Prozorro API. I am ready to take on the project. Write to me, and we will discuss.

Budget: 27000 UAH Deadline: 14 days

Hello, Petro.

I looked at the description. This is not a "Telegram bot"; it is a backend platform with a bot as one of the outputs. The architecture should be laid out immediately for future AI/dashboard/CRM; otherwise, it will need to be rewritten later.

Nuances to consider from the outset:

- The Prozorro API has a lifecycle and versions of lots, deduplication by hash tender_id + version + lot; otherwise, the history will "float."
- Analytics (winners, % decrease, statistics of customers and competitors) through materialized views or batch on cron, not real-time on the main DB.
- PostgreSQL with JSONB for flexibility of Prozorro fields + pgvector in the schema immediately for future AI semantic search by descriptions.
- Filter engine with per-user rules in the DB; otherwise, with multi-tenant, there will be chaos in configurations.
- Polling new tenders through a queue, not cron on bare Python, to withstand rate limits and retries.

Stack: Python 3.12, FastAPI, aiogram with FSM, Celery + Redis, PostgreSQL 15 + pgvector, Next.js 14 (app router, server components, Auth.js, shadcn/ui) for the admin panel, Docker, Hetzner/DO, GitHub Actions for CI.

A realistic budget for this scope: from 45,000 UAH for MVP, timeframe 2 weeks. It did not pass through the limit in the exchange field, so I am clarifying here.

Questions to start:

1. Is the user single or multi-tenant with several users with their own filters?
2. Is a historical bulk import for 1-3 years needed, or only new tenders from the moment of launch?
3. Analysis of affiliation through EDRPOU and repeated connections between customer-participant: for MVP or in subsequent iterations?
4. Do you see the admin panel as a web dashboard, or is the bot itself sufficient at the start? Expand

Budget: 27000 UAH Deadline: 15 days

Hello. I have extensive experience in developing Telegram bots. I am ready to implement it in Python.

Budget: 27000 UAH Deadline: 25 days

Thank you for the description. For the MVP, I would focus not on 25,000 UAH, but on 140,000 UAH and 25 working days, if we are not just creating a Telegram bot, but a backend system with a database, change history, filters, analytics of previous tenders, and preparation for the AI module.

Essentially, I see the stack as follows - Python, FastAPI, PostgreSQL, Telegram Bot API, Docker. For Prozorro, I would also separately include a synchronization module, processing queue, normalization of tenders, deduplication, change log, and notification rules. A budget of 25,000 UAH for such a scope feels too small =/

Look, there’s a nuance here - the analytics of competitors and previous tenders should be designed as a separate data layer, not as several commands in the bot. Otherwise, later the dashboard, CRM logic, tasks, and GPT analysis will have to be redone.

> Question 1 - what data is needed for the MVP - only new tenders by CPV and keywords or also historical loading for 1-3 years for analytics?
> Question 2 - do we define competitors only by the names of participants or is there a need to work with the EDRPOU, winners, customers, and recurring connections?

Examples of similar works from Ingello
> https://business.ingello.com/vorfahr - close in logic to AI automation, data collection, and recommendations
> https://business.ingello.com/fractal - an example of agency automation and multi-step data processing
> https://business.ingello.com/platforma - a strong related example of a corporate platform with roles, data, and development towards CRM

Main profile for FreelanceHunt - https://systems-fl.ingello.com/ua

From you at the start, I need filtering criteria, a list of CPV, keywords, examples of 5-10 relevant tenders, and the desired logic of messages in Telegram. After that, we can properly break down the MVP into stages and not guess on coffee grounds. Expand

Budget: 25000 UAH Deadline: 7 days

Hello, I worked on a public procurement analytics system for TransparencyBot - integrated 15+ API sources, processed over 50,000 tenders monthly with automatic filtering and notifications.

I’m curious if you plan to use real-time monitoring via WebSocket connection to the Prozorro API, or if a polling approach will be sufficient to track changes in tenders?

I suggest we get in touch; I will provide you with free technical consultation and we can outline a development plan + I will tell you about my team! ✨

Budget: 24500 UAH Deadline: 20 days

Hello.

For the first phase, I would set up a separate synchronization module with the Prozorro API: we read the feed `/api/2.5/tenders` incrementally through `next_page.offset`, store `public_modified`/cursor, and retrieve the full tender card with a separate request `/tenders/{id}`.

In the database, it is important to immediately separate tenders, customers, CPV/items, lots, awards, suppliers, contracts, and notification history. This will provide control over duplicates, a history of changes, and a solid foundation for competitor analytics. For competitors, it is better to focus not only on the company name but also on `identifier.id` / EDRPOU, as names may differ.

The Telegram bot in this scheme will serve as an interface for settings and notifications: CPV, keywords, region, amount, customer, deadline, latest found tenders. Notifications should be stored separately so that the same tender is not sent again after a restart or status update.

Technically, I propose Python + FastAPI, PostgreSQL, SQLAlchemy/Alembic, aiogram, Docker. AI/GPT should be integrated in the second phase after a stable base of tenders and documents is established.

Before starting work, I need to clarify:
- whether historical backfill is needed and for what period
- which CPV/keywords are critical for the MVP
- whether we count competitors by names or by EDRPOU
- which tender statuses should be sent to Telegram
- where the system will be deployed.

We can discuss the details in private messages. Expand

Budget: 25000 UAH Deadline: 15 days

I am a backend developer with experience in creating API services, parsing/integrations with external APIs, working with databases, and building data processing systems. I have worked with Python (FastAPI), PostgreSQL, Docker, and integrations with third-party services, including the Telegram Bot API.
I have a good understanding of backend system architecture: building modular services, working with queues/tasks, avoiding data duplication, and structuring data for further analytics.
I also have experience with AI tools (OpenAI API / GPT) for automating analytics and processing text data, so the direction of project development (AI analysis of tenders, documentation, and competitors) is particularly interesting to me.
For the MVP, I see the following implementation logic:
— integration of the Prozorro API + tender collection module
— filtering system (CPV, region, amount, keywords, deadlines)
— Telegram bot with commands and settings
— PostgreSQL for storage and change history
— basic analytics on tenders and participants
— Docker for deployment and environment isolation
I am comfortable with a long-term work format with gradual development of the system from MVP to a full-fledged analytical platform. Expand

Budget: 25000 UAH Deadline: 18 days

Hello!

I read the project carefully — the task is clear and technically interesting.

The stack completely matches mine: Python, aiogram, PostgreSQL, FastAPI, Docker. Prozorro has an open API (api.prozorro.gov.ua) — I have worked with similar integrations, understand the data structure and the pitfalls (pagination, status updates, deduplication).

What I can cover in the MVP:
— Integration with the Prozorro API: fetching tenders, updates, loading details
— Filtering by CPV, keywords, region, amount, customer, deadline
— Telegram bot with commands /start /settings /keywords /cpv /region /last and notifications
— PostgreSQL with protection against duplicates and change history
— Basic analytics: winners, participants, percentage decrease, statistics on customers and competitors
— Docker-compose for deployment

Regarding AI/GPT — I have connected the OpenAI API to bots, for analyzing tender documentation at the next stage this will be implemented through the Assistant API or RAG approach depending on the volume of documents.

Approximately for the MVP:
Cost: we discuss after a short call — it depends on the details regarding analytics and API load.

I am ready to show demo bots and discuss the architecture before starting. Write to me — we will arrange a call. Expand

Budget: 25000 UAH Deadline: 15 days

Good day.

I developed parsers and a lot of analytics for Prozorro. I created this system on Django with notifications in Telegram.

I have examples of work that I can show.

Budget: 25000 UAH Deadline: 14 days

Good day.

I can take on the MVP of such a system: backend on Python/FastAPI, PostgreSQL, Telegram bot, Docker, regular monitoring of the Prozorro API, filters, history storage, and basic analytics.

Regarding relevant experience: I have worked with Telegram bots, backend architecture, Docker, databases, automation of data collection/processing, and notification systems. For half a year, I led a large project. Specifically, with the Prozorro API, a technical research phase will need to be conducted separately, but the task fits well with this stack.

1. MVP: monitoring, filters, database, Telegram notifications, basic analytics.
2. Next phase: AI/GPT, dashboard, CRM logic, tasks, and advanced analytics.

Estimated cost of MVP: 55,000 UAH.
Timeline: 1-2 weeks.

Before starting, we need to clarify: is the system for a single user or multi-user, what period of history are we analyzing, what specific competitor metrics are needed, and does deployment fall within the first phase. Expand

Budget: 27000 UAH Deadline: 18 days

Hello! I have relevant experience - I build production-grade Python/FastAPI systems with Telegram bots, PostgreSQL, AI analytics; recently, I took 3rd place solo out of 731 teams at the AI Agent Olympics Hackathon in Milan (AI Week 2026 - the largest AI event in Europe) for a multi-agent system on Claude Opus 4.7 + GPT-5.5 + Gemini 3.1 Pro. My background in AI includes 4 years of project management, MSc in Strategic PM, PRINCE2 - so I design architecture that is scalable and documented, not just "a script that works on a laptop."

The stack for your case: Python, FastAPI, PostgreSQL, Docker, Prozorro API (I have read their openprocurement docs), python-telegram-bot with webhook on FastAPI, Redis for deduplication and queues, AI layer on Claude Opus 4.7 for documentation analytics in the next stages. Portfolio in the profile.

Estimated cost for MVP: 45,000 UAH, timeline 18 working days. Breakdown: integration of Prozorro + DB + deduplication (5 days) → filtering CPV/keywords/region/amount (3 days) → Telegram bot with all commands (3 days) → analytics of previous tenders (4 days) → tests, documentation, Docker compose, handover (3 days).

Ready to start this week. If you have questions about the architecture or want to discuss the roadmap after MVP - let me know.

Budget: 25000 UAH Deadline: 1 day

Good day, I am writing on behalf of the company Devoxen. We specialize in your task. We have extensive experience in developing scalable backend systems, Telegram bots, AI/GPT integrations, and analytical platforms. Your project fits well with the architecture on Python + FastAPI + PostgreSQL + Docker with separate services for monitoring the Prozorro API, analytics, and Telegram notifications.

We can implement an MVP with the correct architecture for further scaling: AI analysis of documentation, dashboard, CRM logic, affiliate analysis, and automation of business processes. We can also set up task queues, caching, a duplicate prevention system, logging, and a flexible filtering system for tenders and competitors. We also provide a guarantee and support if desired. We can start working on your project immediately after discussing the technical specifications.

I suggest moving to private messages for a more detailed conversation.

Budget: 27000 UAH Deadline: 1 day

Hello! I am an experienced developer, capable of creating bots of any complexity, ranking TOP-4 in the "Bot Development" category. I also develop websites and API documentation in Python, where I rank TOP-2 as well.

Regarding your project, I have some clarifying questions that will affect the assessment of your task and help understand what you want. Please write to me to clarify all the details!

You can check my skills in my resume 👉 Freelancehunt
My works are also published in my portfolio 👉 Freelancehunt

💻 I am also the CEO of a team consisting of a FrontEnd developer and a BackEnd developer! So if necessary, our team of three can organize a turnkey development for you!

I look forward to your response, thank you.

Budget: 25000 UAH Deadline: 30 days

Good day! I have experience with Python, FastAPI, and the Telegram Bot API.
An example of my work is a Telegram Mini App store with a catalog, cart, order processing, and integration with the bot. You can view it in the Bart_org_sp_bot ((demo) in Telegram.
I am currently actively developing in the direction of backend systems and API integrations. I am ready to implement the MVP of the Prozorro system from scratch with the correct architecture for further scaling.
Implementation timeline for the MVP: 4-5 weeks.
Week 1 — integration with the Prozorro API, setting up the PostgreSQL database, protection against duplicates.
Week 2 — filtering by CPV codes, keywords, region, amount, customer, deadline.
Week 3 — Telegram Bot with commands /start /settings /keywords /cpv /region /last, notifications about new tenders.
Weeks 4-5 — analytics of winners, competitors, statistics on customers, delivery and documentation.
Payment in installments after each week. I am ready to start after agreeing on the details.

Budget: 25000 UAH Deadline: 1 day

Good day. I have everything completely ready for you. And the same tech stack that you need. You buy your hosting - I will transfer everything. And everything is ready. I can do it all for almost free. 25,000 UAH is too large a budget for such a volume.

Budget: 25000 UAH Deadline: 25 days

Hello!

The stack is: Python, FastAPI, PostgreSQL, Docker, Aiogram, OpenAI API. Prozorro has an open REST API — I know how to parse it, filter by CPV, region, and keywords with deduplication.

MVP architecture:
— polling/ingestion service → Prozorro API
— filtering + AI summary via GPT
— Telegram bot on Aiogram 3
— PostgreSQL + Docker Compose

Budget: 25000 UAH Deadline: 20 days

Experience working with the ProZorro system for over 9 years. I am ready to offer you the necessary solutions.

Budget: 27000 UAH Deadline: 20 days

Good day. This is a very interesting and large-scale task, and I would like to participate. I have experience in implementing high-load backend systems with integration of the Prozorro API, specifically: development of the BI Prozorro analytical module, using WebSocket for real-time data updates, automation of public data collection and analysis, building complex filtering and notification systems.

I propose the following technology stack:
Backend: Python + Fast API
Database: PostgreSQL + TimeScaleDB
Monitoring: Websocket + REST clients
Task queues: Redis + Celery
Telegram: python-telegram-bot
Architecture: Docker, Docker Compose

I can provide an estimated cost and timeline after discussing the details of the implementation and the functionality of the MVP. Please message me privately, and I will present a detailed technical plan. Expand

Budget: 25000 UAH Deadline: 10 days

Hello!

The project is clear — what is needed is not a bot, but a backend system with analytics and architecture for extensions (AI, dashboard, CRM). This is my area.

The stack I propose:
Python 3.11 + FastAPI, PostgreSQL + SQLAlchemy/Alembic, aiogram 3.x, APScheduler/Celery + Redis for regular polling of Prozorro, Docker Compose.

Experience:
— Backend on FastAPI + PostgreSQL from scratch to production
— Integrations with external APIs, including Prozorro (I understand the specifics: pagination, dateModified cursor, statuses, lots/awards)
— Telegram bots with personal filters and subscriptions
— Docker, asynchronous workers, monitoring
— OpenAI API — for the second stage of the project

What will be in the MVP:
1. Poller with duplicate protection and change history storage
2. Filter engine: CPV, keywords (with morphology), region, amount, customer, deadline — individually for the user
3. Telegram bot: /start, /settings, /keywords, /cpv, /region, /last + inline menu
4. Analytics: winners, participants, % decrease, statistics on customers and competitors
5. Architecture immediately for future modules (AI analysis of documentation, affiliation graph, dashboard) — so as not to rewrite

Cost of MVP: $2250
Timeline: ~10 days
After MVP: ready to lead the project long-term — AI module, dashboard, CRM logic.

I am ready to write and discuss the details. Thank you! Expand

Budget: 27000 UAH Deadline: 5 days

Good day! I present to you a team of professional developers specializing in AI solutions for businesses! We will create a turnkey solution for you 🙌

An experienced team, ready to offer special conditions for you. Send me a personal message, and I will show you our portfolio and website. I guarantee quality and fast work!

Budget: 25000 UAH Deadline: 20 days

Good day, I have been working with the Telegram API for several years and creating parsers of various complexity, both through official APIs and without them. I am familiar with all the desired technologies for several years, and I also have experience in designing large backend architecture that can be scaled in the future. The estimated cost for the MVP is 25000-27000. The time for implementation is 2-3 weeks.

Budget: 25000 UAH Deadline: 15 days

Good day!

I have been working with Python for over 8 years, regularly creating systems that pull data from external APIs, process it, and deliver it through Telegram bots and web interfaces.

The stack I propose:
- Python 3.12, FastAPI, async PostgreSQL (SQLAlchemy + Alembic)
- python-telegram-bot for the Telegram bot
- APScheduler or Celery + Redis for periodic tasks (Prozorro polling) — the choice depends on the scale and infrastructure
- Docker + docker-compose for deployment
- OpenAI / Anthropic API for the AI stage (documentation analysis, affiliation search)
- Pytest + Ty + Ruff for code quality

Architectural points to consider right away for the MVP (to avoid reworking later):
- Separation into services: ingestion (Prozorro polling), processing (filtering/dedup), notification (Telegram), analytics. Using queues — easy to scale.
- Versioning of tender data (change history through a separate table, not overwrite).
- Flexible filter model — user-based, with the ability to add new criteria without core migrations.
- Preparing the database structure for future dashboard analytics (correct indexes, partitioning by date if needed).
- Scheduler layer — through the interface, so the transition between APScheduler and Celery (if scaling is needed) can be done without rewriting business logic.

Estimated for the MVP:
- Timeline: 2–5 weeks provided there is a clear technical specification and access to the test Prozorro API.
- Cost of MVP: 25,000 UAH. Exact estimate — after agreeing on the details of the technical specification.

Questions that will help to assess more accurately:
1. How many users are expected for the MVP? One user (internal tool) or a multi-user service with a subscription?
2. How often should Prozorro be polled — in real-time (minutes) or once an hour/day?
3. Competitor analytics — are these basic metrics (top participants, win-rate) or deeper (connections between companies, patterns)?
4. Hosting — your server, VPS, or cloud (AWS / Hetzner)?

I am ready for a long-term collaboration — judging by the roadmap (AI documentation analysis, dashboard, CRM, affiliation), the project has significantly greater potential than the MVP.
I can lead it completely or in partnership with your team.

I can send my CV with a full list of projects in private. I am waiting for the technical specification or clarifications — I will return with a final estimate. Expand

Budget: 25000 UAH Deadline: 1 day

Hello! I have relevant experience specifically for this project.
Relevant experience:
I developed and maintain an auto-forecasting system with data parsing from various sources, background processing, task queues, and a Telegram bot — which serves both as notifications and error monitoring on the server. I also implemented a personal account with a filtering system, where the user configures what to receive. Architecturally, this is very close to what is needed in your project.
Technology stack:
Python, FastAPI, PostgreSQL, Redis, Celery, Aiogram 3, Docker / Docker Compose, SQLAlchemy + Alembic.
Approach to MVP:
I will lay out the architecture with a view towards future AI analysis, dashboard, and CRM logic — to avoid rewriting from scratch in the next stages.
Estimated cost of MVP: 30,000 UAH (~$720)
Timeline: 4–5 weeks
I am ready to discuss the details and answer any questions.

Budget: 25000 UAH Deadline: 5 days

Good day.

I am responding to the AI monitoring system for Prozorro + Telegram Bot. This is exactly my profile: not a "bot for the sake of a bot," but a scalable backend (API → filters → database → analytics → notifications) with a foundation for AI in the next stages.

Relevant experience
I have implemented and support a production-like system for monitoring external sources + Telegram (Python, aiogram 3, httpx, pydantic, background poller, SQLite/history, anti-duplicates, personal filters per user):

- several data sources in parallel;
- filters: region, amount, key parameters, multiple profiles per user;
- bot commands, FSM, /start, settings, instant check;
- notifications only for new entries;
- resilience to API changes and repeated requests.

Architecturally, this is 1:1 with your MVP for Prozorro, only instead of OLX/DIM.RIA/Realtor — Prozorro Open Data API + PostgreSQL instead of SQLite for scaling.

I am familiar with Prozorro/OpenContracting: tenders, plans, contracts, CPV, regions, updates by dateModified, deduplication by id. I will create an MVP tailored to your filters and analytics for winners/competitors. Expand

Budget: 27000 UAH Deadline: 21 days

Good day. I, an individual developer, offer my services for the design and implementation of an AI-based tender monitoring system, having experience in full-stack development, building backend architecture, and integrating AI/OpenAI services. I work officially through a sole proprietorship. I independently carry out the entire MVP development cycle: from setting up an asynchronous client for the Prozorro API to deploying a PostgreSQL database with unique indexes to protect against duplicates and logging the history of status changes in tenders. The bot on Aiogram 3 will support all necessary filter management commands. Since the project is long-term, the code will be fully typed (Pydantic / TypeScript-ready) for further seamless integration of CRM logic and dashboards. My approach to optimizing the architecture is presented in the portfolio at 3magency.co and https://vilors.studio/. The exact cost and development timeline will be fixed after reviewing the complete functionality checklist.

Budget: 25000 UAH Deadline: 16 days

Hello. Our small team has over 4 years of experience in full-stack development, database integration, and creating intelligent monitoring systems. For the MVP of Prozorro analytics, we propose the stack: FastAPI + PostgreSQL + Redis (for caching the tender feed and bot sessions) + Docker. The keyword and CPV filtering system will be optimized using Full-Text Search (FTS) in Postgres. The competitor analytics module will collect and aggregate the history of disqualifications and victories of companies, providing clean statistics in Telegram notifications. The project is fully ready for further scaling: adding a web panel and connecting LLM models for analyzing PDF/DOCX specifications from the client. You can familiarize yourself with our projects and the quality of our work at the following links: freshagro.com.ua, farfieworldwide.com, rivnekolo.com. A detailed estimate and timeline will be provided after agreeing on the business logic.

Budget: 25000 UAH Deadline: 7 days

Good day. My colleague and I specialize in developing complex backend systems, integrating government APIs, and building analytical solutions in Python (FastAPI / SQLAlchemy / Celery). We see the priority of the first stage as creating a fault-tolerant framework for collecting tenders and tracking their change history. The logic of analytics for previous periods (number of participants, statistics of winners and customers) will be implemented at the level of optimized SQL queries and data warehouses in PostgreSQL, which will lay the correct foundation for future AI analysis of tender documentation (via OpenAI API / LangChain). We have experience in containerizing complex applications using Docker-compose for deployment in CI/CD environments. Examples of our architectural approach to building stable commercial platforms: drkukharevich.rivne.ua, crave-agency.com.ua, jk-solution.com.ua. We will form the estimated budget and timelines based on the results of the discussion of the technical specifications.

Budget: 25000 UAH Deadline: 12 days

To build a scalable monitoring system for Prozorro, my partner (full-stack architect + Python backend engineer) and I propose to design a distributed architecture based on FastAPI and PostgreSQL. We will implement the work with the Prozorro API (through public and qualification endpoints) as a separate asynchronous service-worker for background data streaming, which will protect the system from timeouts and record duplication. The database will contain a normalized table structure for quick analytics (calculating the percentage decrease in price, building a competitor profile by EDRPOU) and flexible filtering by CPV and regions. A Telegram bot on Aiogram 3 will serve solely as a lightweight client interface. Our experience in developing high-load analytical solutions, parsers, and automation systems is reflected in the projects hyperfi.tech, espressolab.com.ua, and hudi.com.ua. We are ready to provide an estimate of costs and timelines for the MVP after agreeing on the relational data schema.

Budget: 25000 UAH Deadline: 5 days

Good day, I am interested in the offer. Please write in private messages to discuss the details. I will be happy to collaborate!

Budget: 25000 UAH Deadline: 5 days

Good day! I already have a ready-made Telegram bot code for Prozorro in Python+PostgreSQL. This is almost your MVP. Write to me privately, and I will give you the link to take a look. I will adjust it to your needs.

Budget: 25000 UAH Deadline: 25 days

Good day. I can write such a bot in Python. Yes - I have already worked with the chatgpt integration. Python + PostgreSQL + Telegram Bot Api.

Budget: 25000 UAH Deadline: 7 days

Good day!

I professionally develop scalable backend systems and Telegram bots in Python (FastAPI, PostgreSQL, Docker). I have experience integrating with the Prozorro API and building analytical solutions. I am ready for a long-term project with future AI implementation.

Message me privately to discuss the details.

Budget: 25000 UAH Deadline: 3 days

Hello!

I already have an active monitoring system for Prozorro for another client. It does most of what you described in the technical specifications, just for a different product segment. This is a working product with real business logic.

What has already been implemented and works:
— Two workflows instead of one: "Leads" (who has already purchased — through the contracts endpoint) + "Active Tenders" (where submissions can be made — through the tenders endpoint). These are different business tasks, and they need to be architecturally separated.
— Filtering by CPV codes, keywords, and brands, minimum amount
— Deduplication by contract and tender identifiers
— Extraction of auction participants — name + EDRPOU + initialBid + finalBid + submission date (for cold outreach on competitors)
— Tender files and proposal register (xlsx)
— Schedule trigger with configurable frequency (4x/day)
— Telegram notifications about new matches
— Purchaser contacts with all edge cases (fallback to delivery address for military units without city/region)

Answers to your questions:

1. Relevant experience:
— Active Prozorro Monitoring system (https://freelancehunt.com/showcase/work/avtomatizovana-lidogeneratsiya-ta-monitoring-tenderiv/2041398.html)
— Multi-channel data pipeline: Prozorro + World Bank API (https://freelancehunt.com/showcase/work/multi-kanalna-sistema-monitoringu-tenderiv/1974821.html)

2. Technology stack:
My current system — n8n (self-hosted) + PostgreSQL/Supabase + Telegram Bot API.
So I propose to build it on n8n — we transfer the existing logic, configure CPV/keywords for your niche, add bot commands (/start, /settings, /keywords, /cpv, /region, /last) and basic analytics.

3. Three working days after access is provided — you will have a working MVP.
n8n self-hosted = a full production stack, my system reliably processes ~1k tenders/day. Expand

Budget: 25000 UAH Deadline: 21 days

Good day. I understood the requirements — a monitoring backend over the Prozorro Open API with filtering by CPV/keywords/region/amount, Telegram notifications with commands for managing filters, a history of tenders with protection against duplicates, and analytics on winners/competitors. The next stage — AI analysis of tender documentation, dashboard, CRM logic.

I see the architecture like this. A separate fetcher worker queries the Prozorro API for lots using cursor pagination, storing two layers in PostgreSQL — raw events plus normalized projections, idempotently through a composite PK with tender id plus revision, to correctly capture tender updates and avoid pushing the same ones again. On top — async FastAPI with endpoints for the bot and future dashboard, a separate aiogram service for Telegram with commands through FSM. We keep user filters in a user_filters table and perform matching on the Postgres side (GIN indexes on CPV array, full-text on keywords). Deduplication of notifications — a separate notifications table to guarantee one push per user_id plus tender_revision.

Analytics — a separate query to Prozorro for the history of the winning tender, calculating the average percentage decrease in price, the number of unique participants, and statistics on DK021 codes. For AI analysis, we can further use embeddings plus RAG on extracts of tender documentation, but this is the second stage after the MVP.

Relevant experience directly related to the stack — I personally developed a voice AI assistant with RAG over Qdrant in production, an AI platform for contextual advertising in Telegram with MTProto, and another Python+Anthropic side project that is currently pulling from several sources plus LLM scoring plus a Telegram bot with JobQueue for commands. Regarding infrastructure — Docker Compose plus Alembic plus pytest plus deep logging are always included.

Approximately for the MVP, without AI analysis of documentation — 3 weeks to a working bot with a full cycle of monitoring plus filters plus notifications plus basic analytics on tenders and customers, plus another week for stabilization and a statistics dashboard in Telegram. AI analysis of documentation plus searching for affiliations plus web dashboard — a separate stage after the acceptance of the MVP.

I am ready to show the GitHub of current projects and discuss the architecture in more detail over voice. Expand

Budget: 25000 UAH Deadline: 10 days

Welcome! We have been professionally engaged in process automation, data parsing, and the implementation of artificial intelligence for over 4 years. Creating an AI system for monitoring and analyzing Prozorro tenders is our specialized stack, where we will apply Python integration, winner prediction algorithms, and processing large data sets. We will design a clean UI/UX interface for analytics, connect stable tender documentation collection, and guarantee absolute security of corporate information. We suggest discussing the logic of AI analysis criteria in the chat to formulate a clear technical specification.

Budget: 25000 UAH Deadline: 10 days

Good day, Petro, I regularly work on projects of a similar nature. My last project is essentially very similar to yours, just with a different theme.
Let's go step by step:
1. To be honest, I haven't worked directly with Prozorro, but I have now familiarized myself with the API. The task is clear, and the API has the necessary methods for its implementation. So I am confident that I can implement it.
2. The database part is clear; I have worked with PostgreSQL many times, so there will be no issues.
3. Filtering and so on can be implemented in various ways, including contextual options (RAG), search using bm25, as well as built-in filters and search methods in PostgreSQL.
4. The backend will be on async fastapi, and I will connect the Telegram bot using aiogram.
5. I am currently working regularly with AI; we can use lammaindex or pydantic AI.

Regarding the implementation, I will build a flexible, adaptive architecture with the ability to scale to multiple streams and lay the foundation for implementing the mechanisms that, as you mentioned, will come after the MVP. I will cover everything with proper tests and conduct quality modular testing of the project. And finally, I will create a Docker container for deployment and use on any server.

I cover your entire stack, and moreover, your task aligns very well with what I often work on. So, I would be happy to work with you and guarantee the quality and stability of your project's implementation! Expand

Budget: 25000 UAH Deadline: 10 days

Мета зрозуміла: масштабований backend з Telegram-ботом і AI-аналітикою, не просто бот. Це наш профіль: Winbix.AI (multi-tenant AI на FastAPI + Postgres, $2K MRR, 30+ клієнтів), BrandSync AI (GPT-4 + Whisper + BullMQ pipeline), BotFusion AI (SaaS чатботів).
Стек MVP:

aiogram 3 + FastAPI + PostgreSQL з SQLAlchemy 2.0
Celery або Dramatiq для polling Prozorro + Redis broker
APScheduler для recurring polling по CPV
Pydantic v2, audit log на історію змін, unique constraints проти дублів
Docker Compose з health-checks
ПостMVP: pgvector для RAG по тендерній документації

Ставка 25к грн — стартовий milestone: paid discovery + Prozorro API integration + DB schema з захистом від дублів, 7-10 днів. Повний MVP розписуємо по фазах і milestone billing, точну вилку дамо в особистих після уточнень по scope. Орієнтовно 3-4 тижні на повний MVP.
Quentar, студія з 6 розробників (UTC+2). На MVP: 1 senior backend на ядро + 1 dev на bot/UI шарі, PM без додаткової вартості.
Портфоліо: winbixai.com/en/startups. GitHub-доступ і Loom-огляд схожого AI-pipeline скинемо в особисті.
Питання:

Single-user MVP чи multi-tenant з власними фільтрами для кожного юзера з самого старту?
Пріоритети по постMVP (RAG, dashboard, CRM) щоб одразу закласти архітектуру? Expand

Budget: 25000 UAH Deadline: 14 days

Hello! I am ready to help with your project. I have extensive experience in development and can implement all necessary components according to your documentation. I guarantee quality execution within the agreed deadlines.

Budget: 24999 UAH Deadline: 30 days

Good day
I possess all the knowledge and technologies to implement such a system, without exception.
Write to me.

Budget: 27000 UAH Deadline: 15 days

Hello, Peter! I am Nina — the manager of the developer Valentin. We have reviewed the technical specifications — the task is excellent.

Valentin understands perfectly that this is not just a "Telegram bot," but a full-fledged data platform (ETL pipeline) for collecting and normalizing data, which needs to be designed for future scalability (AI analysis of documentation and dashboards).

Our approach to the MVP architecture: Valentin will build the system based on an event-driven model. Synchronization with the Prozorro API will be incremental (by dateModified), rather than a full sweep. We will design the database (PostgreSQL) with versioning of lots (through field hashing), so that the history of tender changes does not "drift." The analytics layer (price reduction calculation, competitor statistics by EDRPOU) will be separated to ensure that heavy queries do not overload the main backend and bot.

Stack: Python (FastAPI, asynchronous engine), PostgreSQL (we will immediately lay down structures for pgvector for future semantic AI search in documents), aiogram 3.x, Docker + Docker Compose.

Experience: Valentin has a strong background in designing scalable backends, parsers, and analytical systems with AI agent integration (OpenAI API).

MVP implementation conditions:
Timeline: 14–16 days (in stages: database and Prozorro pipeline -> analytics layer -> bot interface).
Cost: 48,000 UAH.

We have one clarifying question: for competitor and customer analytics at the MVP stage, will we need to extract historical data from Prozorro for past periods (for example, for 1–2 years), or do we start accumulating statistics and analyzing only those tenders that enter the system from the moment of launch?

Please write to us in private messages, Valentin is ready to discuss the database structure in detail and show examples of architectural solutions. Expand

Budget: 27000 UAH Deadline: 10 days

Hello, answers to the questions
- extensive experience with AI agents implementing them in Telegram as well, wrote many backends, and worked with prozoro
- Stack: FastAPI, PostgreSQL, OpenAI
- examples https://zem.center/ (prozoro), https://217-154-170-186.nip.io/ (AI)
- cost $900
- 10 days

My GitHub https://github.com/onyx144

Budget: 25000 UAH Deadline: 20 days

Good day, Petro. The project looks like a full-fledged monitoring and analytics system, not just a Telegram bot — it actually requires a backend service with Prozorro parsing, a database, filters, and subsequent AI analysis.

I work with Python (FastAPI), PostgreSQL, Telegram Bot API, Docker. I have built similar systems for data collection, filtering, task queues, and Telegram notifications, so architecturally this is a clear task.

At the MVP stage, I will implement the connection to the Prozorro API, collection and updating of tenders, basic filtering (CPV, region, amount, keywords), storage in PostgreSQL, and a Telegram bot with commands and notifications. I will also lay out the structure to later add AI analysis, competitive analytics, and a dashboard without restructuring.

Regarding the MVP timeline — approximately 2–4 weeks depending on the details of the filters and load.
Regarding the budget — starting from $700, more accurately after clarifying the scope of analytics and data update frequency.

If there are technical preferences regarding the architecture or examples of similar systems, I can immediately adapt the solution to that.

I suggest discussing the details. Expand

Budget: 25000 UAH Deadline: 30 days

Hello!

I am interested in the project - it is not a "notification bot," but a proper data platform around Prozorro with a clear path to AI analytics. Such systems fit well into an event-driven architecture: collection -> filtering -> analytics -> delivery to Telegram.

Experience and stack: Python, FastAPI, PostgreSQL, aiogram/python-telegram-bot, Docker, Celery/APScheduler for background tasks. I have worked with REST APIs, ETL pipelines, deduplication, and change history of entities - this is exactly what is needed for tenders.

How I would create the MVP:

Prozorro API - incremental synchronization by dateModified, not a full rebuild every time. Tenders, CPV, lots, awards, contracts - in normalized tables with external_id + hash fields to protect against duplicates and versioning changes.

Filters - a separate layer of user rules (CPV, keywords, region, amount, customer, deadline). First, strict filters in the database, then scoring by relevance - to avoid spamming in Telegram.

Telegram Bot - /start, /settings, /keywords, /cpv, /region, /last through inline keyboards and FSM. Notifications with a brief card + link to the tender + "why it was included in the selection."

Analytics - aggregates on winners, number of participants, % price reduction, activity of customers and competitors. This immediately provides value without GPT, and later it makes sense to integrate AI based on documentation.

Architecture for growth - with the MVP, I lay down modularity: ingestion, matching, analytics, notifications, bot - separate services/modules, so that later it is easy to add a dashboard, CRM, and AI.

Similar tasks: monitoring of public procurement/tenders, alert systems based on external APIs, Telegram bots with personal filters and analytics on historical data.

MVP timeline: 3-4 weeks

1-2 weeks: API + DB + synchronization
2 weeks: filters + bot + notifications
1-2 weeks: analytics + tests + deployment
Estimated cost for MVP: (depends on the depth of analytics, number of filters, admin panel, and volume of historical data for initial loading).

I am ready to discuss priorities: what is more important at the start - speed of alerts, depth of competitor analytics, or preparation for the AI stage. I can sketch out the database schema, pipeline, and roadmap after the MVP during a call. Expand

Budget: 25000 UAH Deadline: 20 days

Hello! I am Volodymyr, the founder of the engineering agency Vaysed. Your task of creating an analytical system for monitoring Prozorro tenders perfectly aligns with our key technical stack. We specialize in developing high-load architectures, integrating with complex government and commercial APIs, and deploying autonomous analytical bots, so we are ready to implement this project with future scalability in mind.

The entire system will be built on a reliable backend using Python and FastAPI. To store large arrays of tender data, change history, statistics on customers, and to protect against duplicates, we will deploy a powerful relational database PostgreSQL. Interaction with the Telegram Bot API will be implemented through modern asynchronous frameworks, ensuring instant delivery of notifications about new purchases according to user-configured filters. The entire architecture will be immediately packaged in isolated Docker containers for secure deployment and stable operation on servers.

Integration with the Prozorro API requires uninterrupted operation of background processes for constant data updates and accurate filtering by CPV codes, amounts, or regions. We will configure the system to automatically collect historical analytics on competitors and price reduction percentages. This structured database will become an ideal foundation for the next stages of development when we connect AI for deep analysis of tender documentation and searching for affiliations.

Our team has deep practical experience in creating similar fail-safe data pipelines. We successfully implemented the mechanics of continuous monitoring of external sources without interruptions and instant client notifications in our Lumos project. We refined the complex logic of command processing, managing settings through Telegram, and maintaining session context in the large-scale ecosystem RESTART 2026. And we perfectly configured reliable database integration for building further CRM logic in the Prudbay Clothes Support system.

You can evaluate our engineering approach and familiarize yourself with implemented projects on the official agency website https://vaysed.me/. Message me privately so we can discuss the structure of the analytical module in more detail and promptly start designing the database.

Budget: 25000 UAH Deadline: 10 days

Good day! I have experience in developing backend systems in Python, including parsing, automation, Telegram bots, and integration with external APIs. Your project is interesting precisely because it is not just a Telegram bot, but a full-fledged analytical system with long-term prospects.

For the MVP implementation, I envision an architecture based on Python + FastAPI + PostgreSQL + Docker, with separate services for monitoring the Prozorro API, processing/filtering tenders, analytics, and the Telegram bot. For the bot — Telegram Bot API (or Aiogram), with a filter settings system, history, and protection against duplicates in the database.

In the first phase, I can implement: integration with the Prozorro API, monitoring new tenders, filtering (CPV, keywords, region, amount, deadline, customer), Telegram notifications, bot commands, as well as basic analytics of previous purchases (winners, competition, statistics). The architecture can be designed from the outset with scalability for AI/GPT analytics, dashboard, and CRM logic in the future.

Estimated MVP: 1-2 weeks depending on the detail of the analytics and bot scenarios. I will be able to provide a more accurate cost estimate after a brief discussion of the requirements and priorities for the MVP. I would be happy to discuss the details and propose a technical implementation plan.

Budget: 25000 UAH Deadline: 5 days

Hello! I have been working in automation and developing Telegram bots for over 4 years! I have experience with similar projects.

After discussing the specifications, I will start working.

Budget: 25000 UAH Deadline: 3 days

Hello! I have reviewed the project and am ready to start working. I am confident you will be satisfied with the result.

Budget: 25000 UAH Deadline: 25 days

The system will receive tenders via the Prozorro API, filter them by CPV, keywords, region, amount, and deadline — and send a notification in Telegram before the competitor even opens their browser.

The stack fully matches yours: FastAPI, aiogram, PostgreSQL, asyncio, Docker. I am building the Telegram bot with commands /settings, /keywords, /cpv, /region, /last on aiogram — with FSM for settings and a flexible filtering system at the DB level. I am integrating GPT analytics into the architecture from the start to avoid reworking after the MVP.

I have built similar pipelines: data collection and normalization through external APIs, deduplication, queues on Procrastinate, notifications via Telegram, deployment on Hetzner using docker-compose with GitHub Actions CI. The analytical module — winners, percentage decrease in price, statistics on customers and competitors — I am designing as a separate layer to easily expand after the MVP.

Estimated timeframe for MVP — 3-4 weeks. Write to me — we will discuss the architecture details and a convenient format for starting.

Budget: 25000 UAH Deadline: 16 days

Hello. I am familiar with the Prozorro API - I integrated it for automating procurement monitoring, where the data flow is large and duplicates kill productivity. Your scope is not just a bot, but a full-fledged backend with collection, filtering, and analytics. This is a build, not a configuration.

I created a similar system: a parser for public procurement on FastAPI + PostgreSQL + Telegram, with 40+ filters, a history of tender changes, and deduplication by hash. It has been running for a year and a half, handling ~2k tenders per day.

For the MVP - integration with Prozorro, filtering, a bot with commands, a database with deduplication, basic analytics - approximately 2-2.5 weeks. The AI layer is better to move to the next iteration to avoid diluting the scope.

Question: Will the CPV codes be fixed or will the user add them through the bot? This significantly affects the data model.

Budget: 25000 UAH Deadline: 5 days

Good day!

I already have a working prototype of the Freelancehunt monitoring system with a Telegram bot:
— RSS/API monitoring
— AI scoring
— keyword filtering
— Telegram alerts
— adaptive learning
— SQLite/PostgreSQL ready architecture

I work with Python, FastAPI, Telegram Bot API, and workflow automation.

I can quickly implement an MVP for Prozorro:
— tender monitoring
— filtering
— Telegram notifications
— backend API
— preparation for AI/GPT analytics

I am ready to discuss the architecture and implementation stages.
I will execute quickly and well! Expand

Budget: 27000 UAH Deadline: 28 days

we already have almost a ready foundation of a similar monitoring, analytics, and notification system, it can be quickly adapted for Prozorro and launch the first working phase... I am in touch ))

the benchmark for a complete MVP is from 120,000 UAH and about 4 weeks.

if we stick to a budget of 25,000 UAH, I would recommend starting with a technical prototype - Prozorro API, basic filtering, a simple Telegram bot, and checking the logic on real tenders.

the stack is decent - Python, FastAPI, PostgreSQL, Telegram Bot API, Docker.

in terms of architecture, I would create not just a bot, but a backend with separate modules for data collection, filtering, history storage, analytics, and notification queue.

then AI analysis of documentation, search for affiliations, dashboard, and CRM logic can fit in well - without painful rework if the foundation is laid correctly.

there are clarifications regarding the MVP.

- is monitoring new tenders only needed at the first stage, or should historical tenders be imported for statistics as well?
- should AI/GPT be connected in the MVP for real analysis, or just prepare the structure for it for now?

similar cases from Ingello

- https://business.ingello.com/vorfahr - AI and process automation, close to the logic of the future analytical module
- https://business.ingello.com/fractal - agency and automation approaches, useful for a long-term system with tasks and analytics
- https://business.ingello.com/forma-bpm - business processes, roles, tasks, and systemic logic, similar to further development in CRM

the main profile for FLH - https://systems-fl.ingello.com/ua

We can start with a short design phase - database schema, bot scenarios, filtering rules, MVP plan, and accurate estimation.

this reduces risk, because in tender systems, the devil, as always, lies not in the bot, but in the data and exceptions =)

overall, it's fine, our task here is not just a button in Telegram, but a system that won't fall apart after the first 10,000 tenders. Expand