You will get an AI knowledge base that answers questions from your internal docs
Project details
I build AI knowledge bases that let your team ask plain-English questions and get answers directly from your internal documents — no more digging through folders or pinging colleagues.
At Barclays, I built a RAG platform handling 50K+ enterprise documents and 15K daily LLM queries across 4 business units. Document onboarding dropped from 3 weeks to 3 days. p95 inference latency cut ~60% through batching and semantic caching. I bring the same engineering to your business.
What you get: a working chat interface connected to your docs, deployed in 2 weeks. PDFs, Word files, Google Docs, Confluence pages — whatever format you use. Ask it anything. It answers instantly with source citations.
Tech: Python, FastAPI, LlamaIndex, Pinecone/FAISS, Azure OpenAI or Anthropic Claude API.
Live demo: portfolio-one-topaz-65.vercel.app
At Barclays, I built a RAG platform handling 50K+ enterprise documents and 15K daily LLM queries across 4 business units. Document onboarding dropped from 3 weeks to 3 days. p95 inference latency cut ~60% through batching and semantic caching. I bring the same engineering to your business.
What you get: a working chat interface connected to your docs, deployed in 2 weeks. PDFs, Word files, Google Docs, Confluence pages — whatever format you use. Ask it anything. It answers instantly with source citations.
Tech: Python, FastAPI, LlamaIndex, Pinecone/FAISS, Azure OpenAI or Anthropic Claude API.
Live demo: portfolio-one-topaz-65.vercel.app
AI Algorithms
Transformer ModelAI Applications
Conversational AI, Natural Language Generation, Natural Language UnderstandingAI Development Language
PythonAI Tools
Azure OpenAIAI Models
ChatGPT, GPT-4What's included
| Service Tiers |
Starter
$500
|
Standard
$700
|
Advanced
$1,200
|
|---|---|---|---|
| Delivery Time | 10 days | 14 days | 21 days |
Number of Revisions | 1 | 2 | 3 |
AI Model Integration | - | - | - |
Batch Normalization | - | - | - |
Database Integration | - | - | - |
Detailed Code Comments | - | - | - |
Image Upscaling | - | - | - |
MLOps | - | - | - |
Model Deployment | - | - | - |
Model Documentation | - | - | - |
Model Monitoring | - | - | - |
Model Testing & Optimization | - | - | - |
Model Tuning | - | - | - |
Natural Language Processing | - | - | - |
NLP Tokenization | - | - | - |
Pre-Training | - | - | - |
Prompt Engineering | - | - | - |
Setup File | - | - | - |
Source Code | - | - | - |
About Ankit
RAG Pipeline & LLM Integration Engineer
Gurugram, India - 5:20 pm local time
At Barclays, I built a Generative AI platform serving 50K+ enterprise documents and 15K daily LLM requests across 4 business units. Document onboarding dropped from 3 weeks to 3 days. p95 inference latency cut ~60% through batching and semantic caching.
Live demos: portfolio-one-topaz-65.vercel.app
For freelance clients, I offer 3 focused services:
- AI Knowledge Base Agent ($700 / 2 weeks): ask your internal docs in plain English
- Business Process Automation Agent ($1,400 / 4 weeks): automate your team's manual workflows
- AI Feature Integration ($450 / 1 week): add LLM capabilities to your existing product
Tech stack: Python, FastAPI, LangChain, LlamaIndex, AWS Bedrock, Azure OpenAI, Anthropic Claude API, Pinecone, FAISS, ChromaDB, RAG pipelines, agentic workflows, MCP (Model Context Protocol), vector databases, semantic search, prompt engineering, OpenAI API.
I focus exclusively on AI/LLM engineering. If you need a RAG pipeline, knowledge base agent, or LLM integration -- I can scope and deliver it in days, not months.
Steps for completing your project
After purchasing the project, send requirements so Ankit can start the project.
Delivery time starts when Ankit receives requirements from you.
Ankit works on your project following the steps below.
Revisions may occur after the delivery date.
Documents & requirements intake
You share your files and answer 3 intake questions. I confirm scope and timeline.
RAG pipeline & vector database build
I process your docs, build the embedding pipeline, and configure the LLM + retrieval layer.