Poorak Pasrija

AI & GTM Engineer @ Peoplebox.ai (YC S22) | Machine Learning, LLMs & AI Solutions | Ex-IIT Mandi, NIT Kurukshetra | LangChain, PyTorch, FastAPI

Peoplebox.aiNex-NIT Kurukshetraex-Carbon Crunch (Backed by IIML - EIC)

“

I'm an AI & GTM Engineer at ai (YC S22), where I build production LLM systems that turn ambiguous go-to-market problems into automated pipelines.

More

My recent work spans an end-to-end LLM citation analytics tool (LangChain + OpenAI Responses API + async Python) and a multi-API enrichment system orchestrating io and Ahrefs that has contributed to $255K of Q1 pipeline by replacing ~20 hrs/week of manual My focus is end-to-end, production-ready AI, the path from a model in a notebook to a system serving real users. I've worked across:- GenAI & LLMs - RAG pipelines with reranking, semantic caching, prompt engineering, agentic workflows. My side project NyayaGPT is a deployed RAG system on the Bharatiya Nyaya Sanhita with a 35× latency speedup via Redis semantic - Classical ML & Deep Learning - At NIT Kurukshetra as a Research Intern, I built an audio deepfake detection system combining MFCC + Group Delay features with LSTM, reducing Equal Error Rate from 32% → - Backend & Automation - REST APIs, async Python, Selenium-based agents, and modular service architecture across internships at Open University of Tanzania, Carbon Crunch (IIM-L EIC), and SaiVa - Research - Co-author of "Energy-Efficient Optical Downlink Strategies for Satellite-Based IoT Networks," accepted at SpacSec'25, Russia. Anubhav Research Fellow at IIT Tech I work with daily: Python, PyTorch, LangChain, FastAPI, OpenAI / Gemini APIs, Redis, ChromaDB, Graduating Tech CSE from Manipal University Jaipur in May 2026 (CGPA 68, Dean's List 5th & 6th sem). Open to full-time roles in Applied ML, GenAI, and AI Engineering, especially at AI-first startups where the work goes beyond notebooks into scalable real-world Reach me at com