You will get a production RAG pipeline with RAGAS evaluation and Redis caching
Project details
Most RAG chatbots fail because nobody measures them before they ship. I build retrieval pipelines evaluation-first — every delivery includes a RAGAS report with Faithfulness, Answer Relevancy, Context Precision, and Context Recall scores, so you know exactly what you're getting before it goes live.
My live work: RegulIQ (reguliq.eu) — a 6-agent LangGraph pipeline processing EU regulatory documents in 3 languages on AWS ECS Fargate, with RAGAS Answer Relevancy 0.78. My YouTube RAG chatbot scored Faithfulness 0.85 on clean text. I know what moves these numbers and why.
You get working code, a deployed system, and a documented quality benchmark. If scores don't meet target, I iterate until they do or give you the root cause in writing.
My live work: RegulIQ (reguliq.eu) — a 6-agent LangGraph pipeline processing EU regulatory documents in 3 languages on AWS ECS Fargate, with RAGAS Answer Relevancy 0.78. My YouTube RAG chatbot scored Faithfulness 0.85 on clean text. I know what moves these numbers and why.
You get working code, a deployed system, and a documented quality benchmark. If scores don't meet target, I iterate until they do or give you the root cause in writing.
What's included
| Service Tiers |
Starter
$299
|
Standard
$499
|
Advanced
$899
|
|---|---|---|---|
| Delivery Time | 3 days | 6 days | 12 days |
Number of Revisions | 1 | 2 | 3 |
Model Validation/Testing | |||
Model Documentation | - | ||
Data Source Connectivity | |||
Source Code |
Frequently asked questions
About Rishi Kumar
AI Engineer | LangGraph Multi-Agent Systems | RAG Pipelines | AWS
Trollhaettan, Sweden - 5:57 pm local time
My live work: RegulIQ (reguliq.eu) — a 6-agent LangGraph compliance pipeline deployed on AWS ECS Fargate with CI/CD, pgvector RAG across 3 languages, GPT-4o cross-model critic, and RAGAS evaluation (Answer Relevancy 0.78). I also published context-lens, an open-source LLM diagnostics library with 209 passing tests on PyPI.
What I build for clients:
- Multi-agent pipelines with LangGraph for complex business workflows
- RAG systems that answer accurately from your documents
- FastAPI backends deployable to AWS or your existing infrastructure
- LLM evaluation layers that prove your system works before handoff
I am an MSc AI & Automation student at University West, Sweden, with 2 years of prior industry experience as a Data Scientist. EU-based, available for async work across time zones.
If you need an AI system built correctly, evaluated, and shipped — message me.
Steps for completing your project
After purchasing the project, send requirements so Rishi Kumar can start the project.
Delivery time starts when Rishi Kumar receives requirements from you.
Rishi Kumar works on your project following the steps below.
Revisions may occur after the delivery date.
Define retrieval success criteria and review your document sources
Build ingestion pipeline, embeddings, and vector store