Headline

AI Builder I CSE’26 @ IIITN

About

I build AI systems that work in production - not just in demos. As a CS graduate from IIIT Nagpur (CGPA: 9.28) with a year of hands-on experience building agentic AI at a fast-growing startup, I have architected and shipped orchestration systems, RAG pipelines, and multi-agent workflows that are live and serving real users today. My work spans the full AI feature lifecycle: problem scoping, prompt engineering, LLM evaluation, agent deployment, and iteration. I have worked across model families from Anthropic, OpenAI, and Google -- selecting, benchmarking, and engineering prompts for each based on the task at hand. I use LangChain, LangGraph, and LangSmith for orchestration and observability, and I have built RAG systems with vector database retrieval to give agents dynamic, context-aware decision-making. A few things I have shipped: • 12+ culinary AI agents across 3 product workflows, integrated with vector DB retrieval for real-time SOP selection • Two core orchestration agents (scheduler + backend coordinator) achieving 99% accuracy on AI-evaluator benchmark suites • End-to-end LLM evaluation pipelines and 10+ automation workflows using Python and n8n • A cross-platform expense tracker (SimplySpent) and a BiLSTM emotion analysis model (91.3% accuracy, deployed to production) I sit comfortably at the intersection of product and engineering -- I have written PRDs, run user interviews, managed backlogs, and triaged AI defects, not just written code. Currently seeking full-time roles in AI Engineering, Agentic AI, Prompt Engineering, or Product Management (AI-native products). Let's connect -- or reach out directly at bsanskar123@gmail.com.

Location

Majalgaon

AI stack

anthropic apis

claude code

codex

cursor

gemini

openai

Work experience

AI Product Analyst at Kookar AI, Jun 2025 - May 2026 · 1 yr, Gurugram, Haryana, India, At Kookar AI, I was part of a small, fast-moving team building AI systems that power a real consumer product. My work spanned agent architecture, prompt engineering, LLM evaluation, and product operations -- often all in the same week. On the engineering side, I designed and deployed multi-agent orchestration systems using AWS Lambda and EventBridge, integrating RAG pipelines with vector databases to enable context-aware, runtime decision-making. I worked across model families from Anthropic, OpenAI, and Google -- engineering, benchmarking, and iterating prompts for production reliability. Agents I contributed to achieved 99% accuracy on AI-evaluator benchmark suites. I used LangChain, LangGraph, and LangSmith throughout for orchestration, tracing, and evaluation. I also built 10+ automation workflows in Python and n8n covering metrics tracking, LLM evaluation pipelines, and operational reporting. On the product side, I served as QC and Issue Manager for all AI features -- triaging defects, coordinating with engineering on resolution, and tracking reliability improvements across releases. I contributed to feature prioritization, agent performance metrics, UX reviews, and cross-functional product discussions. Skills applied: Python, LangChain, LangGraph, LangSmith, OpenAI API, Anthropic API, Gemini API, RAG, Vector Databases, AWS Lambda, AWS EventBridge, Prompt Engineering, LLM Evaluation, n8n, Product Management, Agile

Growth Intern at Fact App, Jun 2023 - Jul 2023 · 2 mos, Nagpur, Maharashtra, India, Fact is early-stage non-profit startup.It is an platform dedicated to counselling and fostering networks among students, faculties, and parents.

Education

Bachelor of Technology - BTech, Computer Science, Indian institute of information technology Nagpur

Public links

github: https://github.com/SanskarBajaj123