AI platform for academic authors

Вітаю!
Маємо досвід у Python / FastAPI, OpenAI / Claude integrations, RAG-системах та AI backend automation.

Працювали з:
- FastAPI production backend
- PDF generation (WeasyPrint / ReportLab)
- OpenAI / Claude API
- RAG та vector search
- PostgreSQL
… - email delivery systems
- Linux / VPS deployment

Можемо швидко включитись у ваш workflow та брати milestone-задачі без постійного мікроменеджменту.

Комфортно працюємо з:
- AI endpoints,
- structured JSON outputs,
- PDF reports,
- browsing/RAG logic,
- async backend задачами,
- API integrations.

Будемо раді обговорити деталі співпраці 🙂

Hello, I can connect to isolated milestone tasks in your FastAPI stack.

I have practical experience with backend workflow, LLM endpoint, PDF generation, and delivering reports to users. For your tasks, I see separate blocks: PDF templates via WeasyPrint/ReportLab, retry logic for delivery, auto-cleaning of files, topic generator via LLM, checklist scoring with JSON/PDF report.

I work carefully alongside the lead developer: first, I agree on the API contract, then I create a small working module, tests, edge cases, and a brief documentation for launching.

Show original

Translate

ітаю,

Прочитав детально — Middle Python Backend Dev для milestone-задач M1–M4, з готовим lead developer і архітектурою. Це мій профіль 1:1.

Стек, з яким працюю щоденно: FastAPI + Pydantic v2 + SQLAlchemy/asyncpg + PostgreSQL, LLM-інтеграції з Anthropic (Claude) і OpenAI з prompt caching, RAG поверх pgvector, browsing-агенти через Playwright, deploy на Hetzner + Docker. Я свій side-проєкт тримаю саме на цьому стеку — Claude API + prompt caching + FastAPI + SQLite, з MCP-сервером для управління.

По кожному milestone коротко:
- M1 PDF + email: WeasyPrint надійніше за ReportLab для довгих документів, Resend — гарний вибір. Auto-delete через 24 год — APScheduler або сам Resend (вони підтримують через API). Retry — tenacity з exponential backoff.
- M2 Topic Generator: FastAPI endpoint + Claude streaming, optional browsing через Playwright або через ваш існуючий browsing agent.
… - M3 Article Readiness: structured output (Anthropic tool-use або OpenAI function calling), JSON schema на чек-лісті, PDF звіт через ту ж M1-інфраструктуру.
- M4 Article Brief: browsing agent → витяг aims & scope → RAG retrieval поверх instruction files → LLM generation.

Як працюю: чисто беру милстоун, уточнюю питання якщо щось у ТЗ неоднозначне, доводжу до passing tests + code review від lead, після прийняття переходжу до наступного. Не зникаю, комунікую через будь-який канал (Slack/Discord/Telegram).

Готовий до короткого діалогу і пробного M1 — це найшвидший спосіб перевірити сумісність.

Good day, I have extensive experience with FastAPI and PostgreSQL. I can take a clear specification and deliver the task to completion without micromanagement. I am ready to get started quickly.

Show original

Translate

Good day! My name is Valentin, and I represent Arctic Web Agency. We are a team that specializes in creating modern and effective solutions for businesses. I can provide examples of our similar work in personal messages. We are ready to take your project to work!

Sincerely, Arctic Web Team
Freelancehunt

Show original

Translate

Worked with FastAPI + OpenAI/Claude API on production projects - AI content analysis, automatic report generation, integrations with payment systems.
Relevant experience:

FastAPI backends with LLM calls (prompt engineering, function calling)
PDF generation via WeasyPrint for reports
Email delivery through Resend/SendGrid with retry logic
Browsing agents on Playwright for data extraction
PostgreSQL read-only operations with existing schema

… Milestone format: used to working with isolated tasks, clear specifications, and code review.
Experience with AI tools: OpenAI API, Claude API, basic RAG through pgvector.

Questions:
Is there documentation for the existing codebase or will there be an onboarding call?

Rate: $15/hour or fixed price per milestone depending on the complexity assessment after getting acquainted with the project.

Show original

Translate

Hello!

I am interested in your project. I work with Python and FastAPI in production, and I also have experience integrating LLM (OpenAI / Claude), building backend APIs, and AI logic for application systems.

I have worked on tasks related to LLM text analysis, generating structured results (JSON/PDF), integrating external APIs, email delivery, and automating backend processes. I have also worked with RAG approaches and agent scenarios (including browsing/data extraction logic).

I am comfortable with the stack: FastAPI, PostgreSQL, LLM API, PDF generation, Resend, deployment on Linux.

I am ready to quickly engage in milestone format tasks and work in a team process with code review.
…
I can start with the upcoming tasks.

Show original

Translate

Hello!

Your stack matches our profile: FastAPI, Claude/OpenAI API, Playwright, PostgreSQL. I have experience with each of the milestones you described.

Regarding the tasks:

M1 — PDF + email: I have done PDF generation using ReportLab, delivery with retry logic, auto-cleaning of files — a standard task.

M2/M3 — LLM integration: I have worked with Claude API and OpenAI, JSON-structured output, scoring systems — I understand how to build this correctly.
…
M4 — browsing agent + RAG: I have experience with Playwright for parsing dynamic pages, familiar with the RAG approach.

The format of isolated milestone tasks is convenient — I take the task, submit the result, without unnecessary micromanagement.

I am ready to discuss the details for each!

Show original

Translate

I can connect to your Python/FastAPI backend and quickly integrate into the current team as a lead developer. I have practical experience with backend/API tasks, AI integrations, and code reviews, so I can dive into the existing code without a long ramp-up. Could you please let me know which 1-2 tasks need to be prioritized at this stage?

Show original

Translate

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

Show original

Translate

Hello, I am ready to start immediately, I do not require micromanagement, but I always clarify the nuances that I do not understand first to complete the task.
- Implemented daily report generation for orders in PDF and Excel format.
- Implementation of corporate email and corporate cloud (Nextcloud).
- Telegram bot for orders, with rich functionality, a separate web application for bot administration (distribution of additional products according to orders, route reports, warehouse reports, etc., adding retail points, nomenclature to the bot from a convenient admin panel).
I am ready to start as soon as possible or to complete a test task.

Show original

Translate

Hello! I am Volodymyr, the founder of the engineering agency Vaysed. The AI platform project for academic authors is a great initiative. The specifics of working with scientific articles, strict journal requirements, and indexing in Scopus are very familiar to me due to my ongoing work with academic research for university practical tasks. I fully understand the pains of authors and the importance of precise LLM analysis of annotations.
I am ready to quickly integrate into your team as a strong performer. I am most comfortable with a work format that includes clear isolated milestone tasks under the guidance of your lead developer, where there is already a ready PostgreSQL schema and an architectural foundation laid on FastAPI.
Regarding the main milestones:
For **M1**, I will implement PDF report generation through WeasyPrint (it is more flexible in working with HTML/CSS templates), set up integration with Resend for email delivery with reliable retry logic, and a mechanism for background file deletion after 24 hours.
In **M2 and M3**, I have extensive practical experience with direct calls to OpenAI and Claude API, particularly with setting up strict JSON responses for checklist scoring. We will extract the aims & scope of the journal through Playwright — a powerful tool that I regularly use for creating browsing agents and parsing.
For **M4**, I will set up a pipeline where the browsing agent will retrieve context via URL, and the LLM will work through a RAG architecture to generate a personalized article brief according to your instructional files.
My main working tool is the Linux OS (Fedora), so working with your Hetzner VPS, deployment, and testing will happen completely naturally and without delays. I write clean code that easily passes code review, work autonomously, always communicate honestly about statuses, and do not disappear during work.
You can evaluate my engineering approach and implemented cases on the agency's website: https://vaysed.me/. Message me privately so we can discuss the details of the first milestone (M1) and start working promptly!

Show original

Translate

Hello. I work with FastAPI and Python. I am ready to collaborate. Feel free to contact me.

Show original

Translate

The most important thing here is not to "invent a system from scratch," but to quickly and stably integrate individual AI modules into an already prepared backend ecosystem. This is the format in which I work most effectively: I take isolated tasks, quickly dive into the codebase, and bring the milestone to a production-ready result without delays and constant support. I have practical experience with FastAPI, integrations with OpenAI/Claude API, PDF generation, API orchestration, and backend logic for AI products. I understand the specifics of working with LLM: retries, structured JSON responses, prompt control, stability of results, rate limits, handling edge cases, and optimizing inference costs. I am also well-acquainted with approaches to RAG, asynchronous task processing, and integrations with external services. I write code in a structured manner, with proper decomposition, logging, and clear API logic, so passing code review will not be a problem. If needed, I can provide examples of relevant work upon request.

Work plan:

1. Analyze the current backend structure, PostgreSQL schemas, API flow, and milestone requirements.
2. Implement M1: PDF report generation via WeasyPrint or ReportLab, email delivery via Resend, retry mechanism, cleanup of temporary files.
3. Implement M2: FastAPI endpoint for AI Topic Generator, integration with OpenAI/Claude API, structured output, error handling, and optional parsing via browsing agent.
4. Implement M3: AI analysis of abstract/article readiness, checklist scoring, formation of JSON responses and PDF report.
5. Implement M4: pipeline journal URL → browsing agent → LLM → article brief generation with support for RAG and instruction files.
… 6. Test endpoints, check stability of AI calls, logging, and optimize response flow.
7. Integrate into the team's Git workflow, prepare for code review, and deploy on Linux/Hetzner VPS.

Show original

Translate

Good day.

The task is clear. We need a Middle Python/FastAPI developer for the team to work on isolated milestone tasks involving LLM, PDF generation, browsing agents, and RAG.

The stack is exactly what we use. FastAPI daily, OpenAI and Anthropic in production, Resend for email, PostgreSQL with pgvector for RAG, Playwright for browsing. These are tools we constantly work with, not ones we are learning for the project.

Relevant case: BrandSync AI, a production SaaS with a GPT-4 pipeline through Redis and queues. Architecturally similar to your tasks: AI content analysis, JSON results based on checklists, generation of personalized documents. Another project, Winbix.AI, is an AI agent platform with RAG and browsing.

Regarding milestones:
…
M1 (PDF + email). WeasyPrint for templates with full CSS, Resend with webhook handling for retries, scheduler for auto-deletion.

M2 (Topic Generator). FastAPI endpoint with LLM-call, structured output through function calling, optional browsing via Playwright.

M3 (Article Readiness). LLM analysis with structured output based on a checklist, JSON schema for each item, PDF report.

M4 (Article Brief). Browsing agent via Playwright for the journal, RAG through pgvector, generation of a brief with instructional files as system prompts.

The format of isolated milestones with separate payments is suitable. We work with a clear technical specification, and we accept code reviews from the lead developer calmly, applying feedback quickly.

We are ready to take one milestone as a test. If the collaboration and code quality are satisfactory, we will continue with the rest.

Portfolio: quentar.space/en/startups

I await your message in private.

Show original

Translate

Hello! I have reviewed your project and am ready to start working. I can guarantee excellent results in a short time.

Show original

Translate

Hello. I have thoroughly reviewed the project description and milestones. The format of isolated tasks with a clear technical specification and a well-understood project foundation suits me perfectly. I have strong experience working with the backend on FastAPI and writing clean asynchronous code that will easily pass code review from your lead. I constantly work with OpenAI and Claude API, particularly for complex parsing, generating structured JSON responses, and scoring texts, which perfectly addresses tasks M2 and M3. I am also familiar with implementing PDF generation through WeasyPrint, followed by sending via Resend, and I handle the logic for retries and file cleanup using built-in background tasks in FastAPI without putting extra load on the system. I know how to work with parsers and RAG systems for report generation and understand how to properly feed the model context. Could you please let me know if you are already using a specific vector database for the fourth milestone with RAG, such as pgvector, or will I need to implement this solution myself as part of the task?

Show original

Translate