Tech Lead LLM Web - Freelance

Project overview

Context

You are joining a leader in its field, in charge of one of the group's most strategic technological components. Alongside a team of passionate experts, you will tackle the company's highest technical challenges and work on the most innovative topics. In an international context, your mission is to implement and provide AI capabilities to the group's products used daily by specialized professionals.

As a Tech Lead, you will be at the heart of our AI services: the technical component that connects our language models, our search engines, and our strategic products. Your playground: transforming the way an entire sector works on a daily basis.

Your mission is not to write a simple script, but to design robust APIs capable of orchestrating LLM calls, RAG pipelines, and massive processing for several teams across Europe. You will ensure controlled response times (streaming, low latency) while managing high technical complexity (concurrency, parallelism, scaling, reliability of AI integrations). It also involves setting up a multi-tenant solution (several countries involved) that can be offered to clients in sovereign mode.

Missions

Design & Development of data indexing and retrieval consumption services:

Web APIs & Performance (synchronous / asynchronous):

Technical Leadership & Governance:

Technology Stack

Working Conditions

• Service Design: Define the architecture of APIs and workers, caching strategy, and resilience mechanisms (retries, circuit breakers, model fallbacks).
• LLM Integration: Design inference chains (prompting, function calling, RAG, embeddings, agents) and industrialize their production deployment, challenging choices through POCs and Benchmarks.
• Performance & Scalability: Define how to balance throughput, latency, and cost (synchronous and asynchronous) to serve our production loads.
• API Definition: Create robust and documented APIs (FastAPI) with clear contracts, Pydantic validation, unified authentication, and country context propagation.
• Performance & Asynchronous: Master asyncio, concurrency, and parallelism to optimize throughput, arbitrating between synchronous and asynchronous processing according to constraints.
• Streaming & Real-time: Implement LLM response streaming and fine-grained management of timeouts, backpressure, and request cancellation.
• Observability & Security: Natively integrate telemetry (logs, metrics, traces) and security guardrails (secret management, Zero Trust, token isolation) at the core of the services.
• Inner Source Governance: Be an influential member of the technical community to ensure the consistency and reusability of shared components.
• Support for Value Stream Teams: Help squads (AI assistant, Search, Publishing...) integrate the capabilities provided by the team with minimal friction.
• Craftsmanship Culture: Promote Clean Code, SOLID, and automated testing practices.
• Outreach: Ensure constant monitoring and contribute to the group's technical influence (articles, meetups, open-source).
• Languages: Python 3.11+ (mastery required), Go, Java / Spring Boot.
• Web / APIs: FastAPI, Starlette, Pydantic, Uvicorn (ASGI).
• Asynchronous: asyncio, httpx, concurrency and parallelism (threads, processes, workers).
• LLM & AI: OpenAI / Azure OpenAI, LangChain, Mistral, Bedrock, etc.
• Cloud & Infra: AWS, Azure, Docker, CI/CD, OpenSearch, Weaviate, PostgreSQL, Redis, Celery, Kafka.
• Tests: Pytest, pytest-asyncio.
• Sector: Publishing & Specialized Technology Services | International Context
• Contract Type / Duration: Freelance
• Location / Scope: France (European Scope)
• Domain: Artificial Intelligence & Backend Web Engineering

Project details

• Start: flexible
• Remote: onsite
• Location: Paris, France, FR
• Languages: French