I build backend systems and AI agents that go from "working demo" to "running in production." Most of my recent work sits at the intersection of LLM agents, document processing, and automation — RAG pipelines, LangGraph agents, and data ingestion platforms that other teams rely on daily.
- 🔭 Currently building document-processing and e-invoicing systems at DeepLogic AI
- 🤖 Recent work: a LangGraph voice-agent pipeline handling live phone conversation.
- 🧠 Also shipped: hybrid RAG agents, LLM-based document verification, and text-to-SQL agents
- 🛠️ Tools I reach for daily: Python, Django, FastAPI, LangChain/LangGraph, Docker, and Claude Code
- 🌱 Learning German (aiming for B1) and picking up Kubernetes fundamentals
| Project | What it does |
|---|---|
| analytics-gpt | Natural-language-to-SQL agent — ask questions in plain English, get results from a live database |
| leetcode-telegram-bot | Telegram bot that sends you coding problems on a schedule based on topic, difficulty, and frequency — backed by ~2,000 stored problems |
| questions-gpt | Early exploration into GPT-driven Q&A generation |
- Email: pingabhishek@proton.me
- LinkedIn: linkedin.com/abhishek
- Website: traceabhishek.com