Browse categories
Explore
Fiverr Pro
English
$
USD
I will be your ai engineer to build a custom agent with claude or gpt
Louis L
About this gig
Need a production-grade AI Agent not just a chatbot demo?
I'm Louis, a Senior Engineer who builds AI Agent infrastructure. I shipped DuDuClaw, an open-source multi-runtime Agent platform unifying Claude, Codex, and Gemini (on npm + PyPI). I'll build the same caliber of system for you.
What you get:
- Custom Agent for your workflow (research, ops, support, data extraction)
- Multi-LLM routing across Claude, GPT-4o, and Gemini
- Tool use & function calling done right
- MCP (Model Context Protocol) integration
- RAG with pgvector, Pinecone, or Qdrant
- Error handling, retries, and Prometheus observability
- Production-ready Docker, Cloud Run, or your stack
My edge:
- Real production code not tutorial demos
- ISO 27001 Lead Auditor encrypted secrets, PII redaction
- Open-source maintainer on npm + PyPI (DuDuClaw)
- 6+ years shipping enterprise systems
Tech: Anthropic SDK, OpenAI SDK, LangChain, MCP, Python (FastAPI), Node.js, Rust, PostgreSQL + pgvector.
Before ordering, message me your use case I'll confirm the right package.
Get to know Louis L
Louis L
Senior FullStack AI Agent Engineer
- FromTaiwan
- Member sinceJun 2026
Languages
English, Chinese
Hi, I'm Louis — Senior Full-Stack Engineer, 6+ yrs shipping production systems.
What I bring:
• AI Agent / LLM — built DuDuClaw, a Rust + Python multi-runtime Agent platform unifying Claude, Codex, Gemini. Published on npm & PyPI.
• Web — React 19, Next.js 15, Node.js, NestJS, FastAPI, PostgreSQL.
• Mobile — Flutter (iOS/Android), incl. a 10K+ DAU app.
• Web3 — Led a 5-engineer team on an RWA tokenization platform (Solidity + Gnosis Safe).
• Security — ISO 27001 ISMS Lead Auditor.
Fast, clear comms, audit-ready code. Let's build something great.
My Portfolio
FAQ
Do you sign NDAs before starting?
Yes — I sign NDAs before any code, credentials, or sensitive material is shared. Both standard mutual NDAs and your own template work. All work is delivered via private GitHub repos with revocable access.
Which LLM should I choose for my use case?
Depends on your need. Claude excels at reasoning and long-context tasks. GPT-4o handles general use and tool calling well. Gemini is strong on multimodal. Message me your use case and I'll recommend the best fit before you order.
Who pays for the LLM API costs?
You provide the API keys and cover your own LLM usage (Claude/OpenAI/Gemini). I help estimate monthly costs based on expected traffic and optimize prompts to minimize spend — often cutting tokens by 30-50%.
Who owns the code I receive?
You do. Full IP transfer on final delivery — including source code, documentation, and deployment scripts. I retain rights only to anonymized open-source patterns shared back to my public repos.
Where will the Agent run? Can you deploy it?
Your choice — Docker, AWS, GCP Cloud Run, Azure Functions, or your own server. Add the Cloud Deployment extra and I'll handle setup, monitoring, and CI/CD pipeline so you get a working URL on day one.
How do you handle sensitive data and API keys?
I encrypt stored secrets with AES-256-GCM and apply PII redaction layers — same patterns as my open-source DuDuClaw project. All secret handling is auditable. As an ISO 27001 ISMS Lead Auditor, I treat security as a first-class concern, not an afterthought.
