Budget: 15000 UAH Deadline: 5 days

Welcome! The Business Atlas team is ready to implement your project. I am Oleg, the project manager of the company, and we specialize in building autonomous AI ecosystems and complex automation systems.
Since the stack is being discussed, we propose a significantly faster and more cost-effective approach: to implement a PIM system based on Low-code architecture and AI agents instead of developing a custom backend in Python from scratch. In my experience, there are over 50 successful projects in the UA/EU markets, where delegating processes to visual platforms reduced Time-to-Market significantly without compromising quality.
Our vision of the architecture (Alternative stack)
•API Integration (n8n / Make): Modular scenarios instead of code. Adding a new supplier from 5-7 required can be done in a matter of hours. n8n easily handles large data arrays (JSON/XML) >50,000 products.
•Database: We keep PostgreSQL (with JSONB), or deploy corporate Airtable / Supabase as a quick storage — we have confirmed successful experience implementing n8n + Airtable connections.
•AI Matching Module (Matching Engine): The workflow in n8n is managed by an AI agent (OpenAI/Anthropic). Initially, names are cleaned algorithmically, then the AI compares characteristics, assigns a confidence score, and in case of doubts (75-90%) sends the product for moderation.
•Admin Panel: Quick SPA interface on Budibase / Retool or Airtable components with complex tables, category tree, and manual mapping screen.
•Export and Output API: n8n automatically generates Excel/CSV/XML (for Prom/Rozetka) and acts as a REST API server for integration with your external systems (CRM, BAS, etc.).
Timeline and Cost Estimate for MVP
Stage 1 (Integration of 5-7 APIs): 1 – 1.5 weeks.
Stage 2 (Master Catalog and AI Mapping): 1.5 – 2 weeks (logic, prompting, calibration).
Stage 3 (Admin Panel): 1 week.
Stage 4 (Export and Output API): 1 week.
Total timeline: 4 – 5 weeks (instead of 2-3 months on code).
Estimated cost: $4,500 – $6,500 turnkey (depends on the choice of DB and filters).
The exact budget will be fixed after conducting an expert diagnosis of your current infrastructure.
Are you ready to talk so we can present how Low-code architecture will meet 100% of your requirements in less time and budget? Expand

Budget: 15000 UAH Deadline: 15 days

Good day! I develop in Python, have worked on similar projects with React/Node.js, and am ready to collaborate.

Budget: 15000 UAH Deadline: 1 day

Good day, I have extensive experience with FastAPI, React, and Vue, parsing is not a problem, you can check the results of my work in my profile. I always perform my work at 101% in terms of speed and quality. I look forward to your message, and in the meantime, I will provide brief answers to your questions:
1. Yes, Python 5+ years, Vue/React 3+, parsing - no problem.
2. Yes, I have experience with OpenAI and Anthropic, used in both simple classifications and complex tasks.
3. 3-5 days, 1-3 days, 1-3 days, 1-3 days (I can provide a more accurate estimate once I review the full specifications).
4. In my opinion, for your task, FastAPI+Vue+PostgreSQL+Redis+Anthropic would be the ideal stack.

Price: To be discussed based on the assessment of the specifications.

Budget: 27000 UAH Deadline: 20 days

Hello. I work with React and Django. I am ready to collaborate. Feel free to contact me.

Budget: 20000 UAH Deadline: 30 days

Do any of your 5–7 suppliers have specific rate limits on the number of requests per minute or outdated SOAP/XML APIs that require individual proxy server configurations or cascading uploads of large files, so we can take this into account when designing the Celery queues?

I am ready to demonstrate my architectural approaches, discuss the specifications in detail, and start the project — I look forward to your feedback in private messaging.

Budget: 15000 UAH Deadline: 7 days

Hello.

I am a NodeJS developer. I have experience with OpenAI API. I know React. I am ready to take on the task. Write to me, and we will discuss.

Budget: 360000 UAH Deadline: 30 days

Hello.
Example of a recently completed task
https://freelancehunt.com/project/razrabotka-ai-sistemyi-golosovogo-obzvona-klientov-dlya/1626849.html
The project looks strong and technically interesting. It is not just a catalog, but an internal PIM system with data normalization, a matching engine, AI verification, and a separate moderation interface. I can join the implementation step by step, without overloading the first iteration, but with the correct architecture for scaling.
I work with Python, FastAPI / Django, PostgreSQL, Celery, Redis, React, integrations with third-party APIs, processing non-uniform data, importing JSON/XML/CSV, as well as with AI integrations for classification, normalization, and semi-automatic mapping tasks. For such a product, I see the optimal stack:
FastAPI for API and service layer;
PostgreSQL + JSONB + GIN for flexible characteristics;
Celery + Redis for background import, recalculations, and AI processing;
React for SPA admin panel with tables, filters, and moderation;
OpenAI or Anthropic for verification of questionable matches and managed AI module.
What is important to lay down immediately:

modular architecture of supplier connectors;
stable master catalog - supplier products model;
separate attribute normalization layer;
two-step matching: deterministic + AI verification;
logging of AI decisions and manual moderation;
export and output API as a separate stable contract, not a "side" function.
Preliminary assessment by stages

Stage 1. API integrations + data unification 1-2 weeks 70,000
Stage 2. Master catalog + matching engine + AI 1-2 weeks 120,000 UAH
Stage 3. WEB management interface / moderation / quality control 2 weeks 120,000 UAH
Stage 4. Export + output API 1 week 50,000 UAH
Overall
implementation period: 4-6 weeks
preliminary budget: 285,000-480,000 UAH Expand

Budget: 120000 UAH Deadline: 14 days

The declared 15,000 UAH will not be enough for the entire PIM system, but for a proper first stage - technical design, prototype of key screens, data model, verification of 1-2 supplier APIs, and a precise cost breakdown. The full development of such a system, based on my feelings, will be closer to 450,000-900,000 UAH and 8-14 weeks, depending on the quality of the APIs, the number of categories, and mapping rules.

I would orient myself on the stages as follows:
> Stage 1 - Supplier APIs, normalization, PostgreSQL JSONB - 3-4 weeks
> Stage 2 - Master catalog, algorithmic and AI mapping - 3-5 weeks
> Stage 3 - Admin panel, moderation, dynamic characteristics, quality of cards - 4-5 weeks
> Stage 4 - Excel, CSV, XML, YML, output REST API, documentation - 2-3 weeks

We have experience with Python, React, complex APIs, catalogs, marketplaces, and tasks where AI not only writes text but makes decisions based on data. For your case, I would not start with large development without a technical stage - otherwise, product mapping will turn into a black box, and then everyone will be heroically catching errors in 50,000 items =)

Similar examples:
> https://business.ingello.com/prime-eva - automation for retail and product processes
> https://business.ingello.com/carveli - e-commerce and systematic work with the catalog
> https://business.ingello.com/vorfahr - AI automation and data processing logic

Main profile of Ingello for the exchange:
> https://systems-fl.ingello.com/ua

For an accurate estimate, two things are needed:
> Provide examples of JSON or XML from 2-3 suppliers
> Indicate whether there is already a reference catalog with categories, SKUs, EANs, characteristics, and mandatory fields

The first stage for 120,000 UAH and 14 days will provide the architecture, database schema, description of mapping rules, a prototype of the admin panel, and a development plan by sprints. After that, we can calculate the full system properly, with risks and boundaries of responsibility, instead of guessing on coffee grounds. Expand

Budget: 26000 UAH Deadline: 15 days

Good day. My name is Dmytro. I have experience in developing complex systems using Python, PostgreSQL, and React, including integrations with supplier APIs, catalog automation, AI data processing, and building internal PIM/ERP solutions.
I have worked with OpenAI and Anthropic APIs for tasks such as classification, data normalization, matching, generating structured responses, and automating business processes.
According to your specifications, I can implement a modular architecture of connectors for suppliers, a master catalog on PostgreSQL (JSONB), processing queues using Celery + Redis, an AI module for product mapping, and a modern admin panel for moderation and quality control of data.
I initially see a stack of FastAPI + PostgreSQL + Redis + Celery + React. This approach scales well and will allow seamless operation with catalogs of over 50,000 products.
I can also implement vector search through embeddings, a confidence score system, manual moderation of questionable matches, XML/YML/Excel export, and a REST API for integrations with CRM, WooCommerce, or BAS.
I will be able to provide an estimate on stages, timelines, and budget after reviewing the supplier API formats and additional project details.
I am ready to discuss the architecture and propose an optimal solution for further scaling.

Budget: 16000 UAH Deadline: 20 days

Good day, I am ready to implement your project.

Experience with Python/React and APIs:

Backend: I regularly develop high-performance APIs using FastAPI. I have a strong understanding of how to optimize PostgreSQL operations using asynchronous ORM (SQLAlchemy/Tortoise), build effective GIN indexes for JSONB, and optimize complex analytical queries. For background processing of 50k+ products, I will set up Celery + Redis with queue separation (one for fast parsing, another for "heavy" requests to LLM).

Frontend: I create SPAs using React (Vite/Next.js) with modern UI libraries (such as Shadcn/ui or Tailwind). Working with complex large tables, dynamically generating forms based on metadata from JSONB, and interactive category trees—all of this will be implemented with rendering optimization (without interface lags during bulk operations).

Experience integrating LLM for mapping and data classification:

I have practical experience with OpenAI API / Anthropic API for similar tasks. To optimize cost and speed, it is best to use the two-step scheme you described: first, we calculate the cosine similarity of vector embeddings of product names locally (using text-embedding-3-small or local HuggingFace models), and we only refer to more expensive models like gpt-4o-mini or claude-3-haiku through a strict system prompt (with JSON-mode response) in disputed cases.

Vision for the technology stack:

Your stack is ideal. It should not be changed. FastAPI will provide asynchronicity and speed, while PostgreSQL with JSONB will offer the flexibility of NoSQL solutions without losing relational ties (which is critical for the "Master Catalog -> Supplier Product" structure). Expand

Budget: 15000 UAH Deadline: 15 days

Hello.

I have relevant experience with Python backend, API integrations, data normalization, catalogs/table structures, and LLM tasks where the model is used for classification, mapping, and verifying ambiguous matches. I can provide proof of experience with similar tasks in personal messages.

For the stack for this project, I suggest: FastAPI, PostgreSQL JSONB + indexes, Celery/Redis for background processing, React for the admin panel, OpenAI/Anthropic for embeddings and LLM verification. For the matching engine, it’s better to immediately lay out a hybrid: SKU/EAN + normalized names + embeddings, and use LLM only for the "gray area" to avoid making the system expensive and unpredictable.

For a budget of 15,000 UAH, I propose to complete the first technical stage:

1-2 supplier connectors;
basic schema master catalog ↔ supplier products;
normalization of SKU/EAN/names/prices/characteristics;
first algorithmic matching;
minimal API or simple results verification screen.
Estimation of the full project by stages:

API + data unification: 2-3 weeks.
Matching Engine + AI: 2-3 weeks.
Admin panel: 2-3 weeks.
Export + REST API: 1-2 weeks.
I suggest discussing the overall budget for the complete system after agreeing on the project base in personal messages, when the API formats of the suppliers and data examples are available.

Before starting, I need to clarify:

if there is a ready structure for the master catalog;
which 1-2 suppliers are included in the first stage;
whether EAN/SKU are available for most products;
which format of the first result is more important: backend/API or minimal admin panel. Expand

Budget: 15000 UAH Deadline: 7 days

Good day!

I have reviewed the specifications. The task is clear: modular connectors for suppliers, a unified master catalog on PostgreSQL (JSONB), two-level matching (rules + embeddings/LLM), an admin panel for moderation and export/API. This is a typical data-heavy B2B PIM, and the approach in the specifications (Strategy/Factory, Celery, a separate screen for tuning AI) is correct.

Experience relevant to the project:
Python and API integrations: development of services on FastAPI with background tasks, parsing/normalization of various formats (JSON/XML), idempotent sync, retry, rate limits, error logging by providers. I have experience building isolated connectors (one module = one supplier) without changing the core — specifically for your Strategy/Factory pattern.

React / complex tables: SPA with ag-grid tables, filters, category tree, bulk operations, moderation screens (side-by-side comparison of supplier cards and benchmarks).

LLM for mapping: integration of OpenAI / Anthropic not only for chat but for classification and verification of matches — embeddings + cosine similarity, confidence thresholds, fallback to LLM in "gray areas" (75–90%), decision log (prompt, score, verdict) for the administrator.

Budget: 15000 UAH Deadline: 1 day

Good day, I have similar work experience and can complete it quickly and efficiently, write to discuss the details.

Budget: 27000 UAH Deadline: 3 days

Good day. I am ready to complete this project as I have extensive experience in application development.

Budget: 20000 UAH Deadline: 14 days

Good day, I have experience with almost everything mentioned in the task. I have worked with OpenAI for classifying conversations between call center managers and clients, assessing how well they meet the requirements. The stack can be as mentioned in the task, or it can be a backend on Node.js.

Deadlines and costs:
1. Integration of 5-7 APIs + unification - 10,000 / 7 days
2. AI module - 5,000 / 3 days
3. Web interface - 3,000 / 3 days
4. Output API - 3,000 / 3 days

Budget: 15000 UAH Deadline: 30 days

Hello! I can complete your project. I have experience. Write to me and we will agree.

Budget: 30000 UAH Deadline: 1 day

Hello! I have carefully reviewed the specifications — the project is interesting and technically clear, I am ready to take it on.

My experience:
I work with Python at a mid-level — I develop backend applications, integrate external APIs, and build background data processing with Celery + Redis. I have practical experience with React: I have developed web applications, including SPAs with complex interfaces.

Regarding LLM — I am currently actively developing a backend application with integration of Anthropic and OpenAI APIs: I have connected endpoints and built the logic for processing model responses. I understand vector embeddings and working with confidence scores for classification tasks.

I have experience parsing various API formats (JSON/XML). I do not have experience with ready-made PIM systems, but the architecture described in the specifications (master catalog, JSONB characteristics, two-step mapping) is clear to me, and I understand how to build it correctly from scratch.

My estimates by stages:
— Stage 1 (API integration, DB, unification): ~2.5 weeks
— Stage 2 (Matching Engine + AI module): ~3 weeks
— Stage 3 (Admin panel, UI): ~2.5 weeks
— Stage 4 (Export, output API, Swagger): ~1.5 weeks

Overall: ~10 weeks (2.5 months considering testing and adjustments between stages).

Cost: 65,000–80,000 UAH depending on the final volume.

I am ready to discuss the details, clarify the stack, or break it down into sprints in a way that is convenient for you. Write to me — we will agree. Expand

Budget: 15000 UAH Deadline: 30 days

Hello
The task is clear. I have experience in Python/React, parsing complex APIs, and integrating LLM (OpenAI/Anthropic) for data mapping and classification.

I would propose the following stack: FastAPI + PostgreSQL (JSONB + pgvector) + Celery/Redis + React. I will create supplier connectors using the Strategy pattern — new APIs will be added as separate modules without changing the core. The mapping is hybrid: first an algorithm (EAN/SKU + normalized names), then embeddings, and questionable matches (75–90%) will be verified through LLM.

Regarding timelines, I suggest an honest approach: the full scope (all 4 stages, 5–7 suppliers, processing 50k+ products) will take about 2–3 months. However, I can deliver a working MVP in just 2 weeks: 2 connectors, a basic database schema, algorithmic mapping + initial embeddings, and a minimal admin panel with manual mapping. After that, I will expand iteratively through the stages.

This way, you will quickly see results and the approach in action, without the risk of missed deadlines. I am ready to discuss details and budget either by voice or in chat.

Budget: 14250 UAH Deadline: 5 days

The project requires the development of a modular system for integrating supplier APIs and unifying the received data — I am ready to create such architecture using Python and FastAPI. The first step is to develop an MVP in a staging environment within 10 days to verify functionality.