Agentic RAG Engineer – AI Education Project | Short-Term Remote

Lebjawi Tech is looking for an experienced Agentic RAG Engineer to help us complete an AI tutoring platform built for education.

This is not a project starting from scratch. The application and RAG architecture are already well advanced and approaching completion.

We need someone who can quickly understand the existing system, work with me through a technical handover, identify what remains, and help bring the project across the finish line.

🎯 What You'll Work On

Take over and understand an existing Agentic RAG architecture

Help complete and improve our AI tutoring workflow

Work with LiteLLM and our existing LLM infrastructure

Improve retrieval quality, agent behavior, and response reliability

Analyze and optimize token consumption and LLM costs

Debug existing RAG/agent issues rather than rebuilding everything unnecessarily

Help prepare the system for production

Document important architectural decisions and remaining technical considerations

🧠 What We're Looking For

Priority will be given to candidates with:

Strong hands-on experience building Agentic RAG systems

Previous experience with education, EdTech, AI tutoring, or educational RAG

Experience with LiteLLM

Strong understanding of embeddings, retrieval, chunking, reranking, context management, and agent workflows

Experience optimizing token usage, latency, and LLM API costs

Ability to understand an existing codebase instead of immediately proposing a complete rewrite

Good written and spoken English

Ability to work independently and communicate clearly when blocked

Experience with production LLM systems is considerably more important to us than certificates.

💰 Compensation & Duration

This is a short-term remote project with a maximum expected duration of two months.

Month 1 — Paid Trial: $600

The first month will be used for onboarding, understanding the existing architecture, working through the handover with me, and demonstrating your ability to contribute effectively.

If the collaboration is not working, Lebjawi Tech may end the engagement during the trial period in accordance with the agreed terms.

Month 2 — Negotiated Rate

If the first month goes well and we d ---------- to continue, we will negotiate a higher rate for the second month based on your contribution, capability, and the remaining work required to complete the project.

There is no guarantee that the engagement will require the full two months. Our objective is to successfully finish the project—not artificially extend the contract.

📩 How to Apply

Please keep your application concise and answer the following:

Describe one Agentic RAG system you personally built or significantly contributed to. What did the agents actually do?

Have you worked on an education/EdTech or AI tutoring project? Briefly explain.

Have you used LiteLLM in a real project? If yes, what did you use it for?

Give one example of something you've done to reduce token consumption or LLM costs without significantly reducing answer quality.

Share a GitHub repository, portfolio, demo, or other work sample if available.

Are you comfortable with the $600 first-month paid trial and a potentially higher negotiated rate for month two?

🍌 Attention check: Start the first sentence of your application with the word "BANANA."

Applications that don't follow this instruction may not be reviewed.