Thiga

AI Engineer

Thiga

full-time

Posted on:

Location Type: Hybrid

Location: Paris • France

Visit company website

Explore more

✨ AI Apply
Apply

Tech Stack

AWSAzureNoSQLPythonReactSQLTypeScript

About the role

  • Architect Agent Systems: Design agents using tools such as LangGraph or AutoGen.
  • Inference Optimization: Experiment with latency/cost/quality trade-offs (quantization, semantic caching, selecting SLM vs LLM models).
  • RAG: Implement Retrieval-Augmented Generation approaches.
  • Fine-tuning & Local LLMs: Explore self-hosting (vLLM, Ollama) and fine-tuning models on specific datasets.
  • Industrialization (LLMOps): Set up robust evaluation pipelines (LLM-as-a-judge, Ragas, DeepEval) to ensure non-regression and factuality.
  • Full-stack AI integration: Connect the "brains" (LLMs) to data (APIs, SQL/NoSQL databases, vector stores) via an agent-based approach.
  • Architecture audit: Analyze clients' agentic architecture implementations for optimization or redesign.

Requirements

  • A passionate Software Engineer with at least 2 years of experience: You are proficient in Python and/or TypeScript. 'Spaghetti' notebook code is not sufficient.
  • AI Stack expert: You have hands-on experience with agentic frameworks (LangChain, LangGraph, LlamaIndex, or others) and experience with one or more vector stores (Qdrant, Weaviate, Pinecone, ...).
  • Pragmatic and rigorous: You know how to manage hallucination and nondeterminism. You understand the difference between a ReAct agent and chain-of-thought.
  • DevOps-aware: You are comfortable deploying agents on one or more platforms (AWS, Google Cloud, Azure, Hugging Face, ...).
Benefits
  • An opportunity to build autonomous systems, not just chatbots.
  • A technical environment free of legacy AI technical debt.
  • A team that values hard tech and clean software engineering.
  • The opportunity to define AI design patterns for years to come.
Applicant Tracking System Keywords

Tip: use these terms in your resume and cover letter to boost ATS matches.

Hard Skills & Tools
PythonTypeScriptLangGraphAutoGenLLMvLLMOllamaQdrantWeaviatePinecone
Soft Skills
pragmaticrigorouspassionatemanage hallucinationmanage nondeterminism