Freelancer profile translated to English.
Back to original languageAbout Tom
Hi, it's Tom đź‘‹
I develop AI agents and multi-step reasoning systems with LangGraph, LangChain, and cloud infrastructure for enterprises and startups.
→ Multi-Step Reasoning Agents with LangGraph (decision trees, orchestration)
→ RAG Systems & Document Intelligence (vector search, embeddings, precise citations)
→ Cloud-Native AI Deployment (AWS Lambda, GCP Cloud Functions, scalable backends)
Looking forward to connecting!
---
Core AI Agent Capabilities:
🧠Multi-Step Reasoning Agents
I develop LangGraph agents capable of reasoning over previous steps, like my "DeepSearch" algorithm which performs iterative web searches, analyzes if the results are sufficient, and continues until providing accurate answers.
đź“š RAG Document Agents
I create specialized chatbots that analyze complex documents (US budget data), extract relevant chunks, and provide precisely sourced answers.
🔄 Orchestration & Monitoring
With LangSmith and LangGraph, I implement full observability - you see every decision node, reasoning step, and workflow execution in real-time.
Technical Infrastructure:
- Python AI Stack: LangGraph, LangChain, OpenAI API, vector databases (Chroma, Pinecone)
- Cloud Deployment: AWS Lambda, GCP Cloud Functions, serverless architecture
- Data Management: S3/GCS Data lakes, Firestore, DynamoDB, vector search
- APIs: FastAPI, Flask, secure REST endpoints
Why Choose Me:
âś… Proven Track Record
âś… Full-Stack Expertise
âś… Production Focus
âś… Observability
âś… Rapid Prototyping
Ready to build your next AI agent? Let's discuss your needs and see how multi-step reasoning can solve your challenges.
Tom
P.S. If your use case is unique, let's talk. I love new challenges!
French
Native or bilingual
English
Native or bilingual
Russian
Conversational
Can work on-site
Paris (up to 30km)
Experience
- CustomGPT.aiLLM EngineerTELECOMMUNICATIONSJune 2025 - Today (1 year)Paris, FranceContext: Developing the "Explainability" feature for CustomGPT.AI's RAGs. US startup featured in Pinecone's blog as a pioneer in RAG-as-a-Service, with over 400 million vectors hosted on the platform.Tasks:- Develop a complete RAG architecture with Pinecone, OpenAI API, and FastAPI- Conceptualize an LLM-as-a-Judge (claimify) capable of extracting claims from an LLM's output and determining its quality against its context- Implement a Meta-LLM-as-a-Judge capable of evaluating the LLM-as-a-Judge's result based on different stakeholders (legal, security, compliance, accuracy, hallucination, etc)
- TotalEnergiesPython Developer - AWS, CI/CDENERGY AND UTILITIESMarch 2025 - June 2025 (3 months)Paris, FranceContext: Developing the backend for the Data Marketplace at TotalEnergies in collaboration with AWS consultants to ensure scalability and compliance. The Data Marketplace allows Data Owners to securely share Data Products, and users to subscribe to, access, and integrate them.Tasks:- Serverless backend using AWS Lambda (Python) and API Gateway to expose secure REST APIs, designed alongside AWS experts.- Fine-grained access control through IAM roles, custom Cognito authorizers, and API Gateway policies.- Implementation of the subscription workflow, including request processing, approval logic, and data access provisioning.- CI/CD pipelines with GitHub Actions, integration of quality gates using SonarQube- Agile delivery of features, working in sprints with regular planning sessions.Result:A production backend enabling secure Data Product sharingRobust access control compliant with governanceGradual knowledge transfer from AWS consultants to enable team independence
- fieldglass.aiPython Developer - GCP, LLMTELECOMMUNICATIONSFebruary 2024 - March 2025 (1 year and 1 month)Paris, FranceContext: Developing the Python backend for the fieldglass.ai platform with the goal of automatically updating news and receiving and responding to user requests from the frontend (API).Tasks:- Development of a robust scraper and an AI Agent (OpenAI, LangChain / LangGraph, Asyncio integration) for automatic content generation and summarization.- Exposure of secure asynchronous RESTful APIs (FastAPI) and deployment on GCP Cloud Functions.- Synchronization and automatic updating of the Data Lake / Cloud Storage based on the agent's feedback.- Application of Agile/SAFe methodologies, Git, and code reviews.Result:Automation of real-time news updates on the platform.Successful integration with the OpenAI API enabling content generation
Recommendations
Be the first to recommend Tom
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Engineer - specialization Artificial IntelligenceCY Tech2021Cycle Ingénieur et Prépa à CYTech (Spécialisation en Intelligence Artificielle) Exemples de modules : Natural Language Processing, Computer Vision and Deep Learning, Text Processing