AI Agent Developer | I build custom RAG chatbots & LangGraph multi-agent systems for startups — production-ready, not just prototypes.
Hi, I'm Yeasin — an AI Agent Developer specializing in production-ready conversational AI systems.
I build custom RAG (Retrieval-Augmented Generation) chatbots that answer questions using your own documents, policies, and data — not generic pre-trained knowledge that hallucinates or gives wrong answers. I also design multi-agent systems using LangGraph, where different AI "agents" handle different parts of a workflow (diagnosis, drafting, review, approval) instead of relying on one fragile prompt.
What I've built:
- Vendra — a multi-agent conversational e-commerce assistant with human-in-the-loop refund workflows, JWT authentication, Redis caching, and a Next.js frontend. Built with an adapter pattern so a single codebase can serve multiple client stores.
- AmzGuard AI — a 6-node multi-agent system that helps Amazon sellers appeal account suspensions, using RAG over real policy documents so every appeal is grounded in actual policy language, not guesses.
- Jarvis — a fully offline, local voice assistant with wake-word detection, speech-to-text, text-to-speech, and a hybrid AI router.
- DocuMind — a RAG-based PDF chatbot with parent-child retrieval and hybrid semantic + keyword search.
My stack: Python, LangGraph, LangChain, FastAPI, PostgreSQL, Redis, Next.js, and OpenAI/Claude/Groq APIs.
I'm new to Fiverr, so I don't have reviews yet — but I have real, working products I can show you, and I'm pricing accordingly while I build trust here. I care more about delivering something that actually works correctly than just checking boxes on a requirement list. I test thoroughly, communicate clearly, and I don't disappear mid-project.
If you need a chatbot that actually understands your business — not a generic template — let's talk about your project.
Work Terms
How I work:
- Communication: I respond within 24 hours, usually much faster. I'll keep you updated at each major milestone rather than going silent until delivery.
- Requirements: Before starting, I'll ask specific questions about your use case, your data/documents (for RAG projects), and your existing tech stack (if any). The more detail you give upfront, the faster and more accurate the build will be.
- Revisions: I include revisions as specified in the package, but I also do my own internal testing before delivery — so most deliveries are already close to final on the first pass.
- Scope changes: If requirements change mid-project (happens often with AI/chatbot work), I'll let you know clearly if it affects timeline or price before proceeding — no surprise charges.
- Source code & documentation: You get full source code ownership and a brief setup/usage guide with every delivery, not just a black-box demo.
- What I need from you: Access to relevant documents/FAQs/policies (for RAG projects), API keys if you want a specific LLM provider used, and timely feedback during review stages so we don't lose momentum.
- Timeline: I'll give you a realistic estimate based on project complexity before we start — I'd rather be honest about timeline than overpromise and deliver late.
Let's build something that actually works for your business.
Attachments (Click to Preview)
-