Pfizer logoCurrently an Extern @Pfizer·🟢 Open to work
Laxman Kumar Busetty portrait

Machine Learning @ Vosyn

I build AI projects that turn unstructured documents into structured, searchable data using OCR, extraction, and RAG pipelines.

Work samples

Projects focus on AI-powered document insights: converting mortgage and other unstructured files into structured data via OCR, extraction, and retrieval-augmented generation.

  • Outamation Advanced AI-Powered Document Insights and Data Extraction Externship
    Outamation Advanced AI-Powered Document Insights and Data Extraction Externship
    Outamation logo

    Outamation · Jan 2026 · ✅ Verified by Extern

    Outamation Advanced AI-Powered Document Insights and Data Extraction Externship

    An Outamation externship on AI-powered document intelligence, focused on turning unstructured mortgage files into structured, searchable data using OCR, extraction, and RAG pipelines.

    Data AnalysisAI & MLGoogle ColabLlama IndexPresentation SkillsPython
  • FounderOS - AI-Powered Chief of Staff
    FounderOS - AI-Powered Chief of Staff

    Solo-built a multi-agent system at the Agentic Dev Tools Hackathon (Entrepreneur First, San Francisco), orchestrating 4 specialized AI agents Planning, Market Intelligence, Opportunity Discovery, and Executive Briefing to take a founder from a single goal to a full execution roadmap. Generated complete, structured roadmaps in under 10–15 seconds end-to-end; validated across 3+ real startup use cases. Designed structured output validation between each agent layer.

    Next.jsTypeScriptInsForgeMulti-Agent Orchestration

About me

I build AI projects that turn unstructured documents into structured, searchable data using OCR, extraction, and RAG pipelines.

I am an AI/ML Engineer dedicated to building production-ready AI agents, LLM systems, and scalable ML pipelines for customer and enterprise applications. I excel in Python, FastAPI, LangChain, LlamaIndex, and hybrid retrieval. My experience includes LLM evaluation, safety guardrails, and cloud deployment. I have designed, deployed, and improved reliable AI systems in document intelligence, conversational AI, and voice workflows, all with a strong builder mindset that thrives in fast-paced enviro

Externships

Pfizer Advanced AI-Powered Document Intelligence Externship

Pfizer

Experience

Machine Learning Engineer Intern

Vosyn Inc · Feb 2026 – Present

AI Engineer Intern (Part-Time)

Tilli · Oct 2025 – Feb 2026

AI-Powered Document Insights Extern

Outamation (via Extern) · Oct 2025 – Dec 2025

AI Research Assistant (Part time) Adversarial Machine Learning

CSUSB · Oct 2024 – Dec 2025

Education

California State University San Bernardino

Masters · Class of 2025

Skills

OCRData extractionRetrieval-augmented generation (RAG)PythonMachine learning fundamentalsNatural language processing

✅ Verified by Extern

Outamation Advanced AI-Powered Document Insights and Data Extraction Externship

An Outamation externship on AI-powered document intelligence, focused on turning unstructured mortgage files into structured, searchable data using OCR, extraction, and RAG pipelines.

Data AnalysisAI & MLGoogle ColabLlama IndexPresentation SkillsPython

Overview

The work produced an end-to-end document intelligence pipeline for mortgage files. It combined OCR and layout-aware extraction, Python-based parsing, and retrieval-augmented generation with indexed embeddings. Deliverables included extraction scripts, OCR comparisons, a Gradio demo, and evaluation metrics such as recall@K, MRR, QA accuracy, and latency measurements.

Outamation Advanced AI-Powered Document Insights and Data Extraction Externship

What I've accomplished

I delivered a complete RAG pipeline and demo that extracted and indexed mortgage PDFs, produced grounded answers with citations, and reported retrieval and QA metrics including recall@K, MRR, accuracy, and sub-2s latency.

Project breakdown

View all work

FounderOS - AI-Powered Chief of Staff

Solo-built a multi-agent system at the Agentic Dev Tools Hackathon (Entrepreneur First, San Francisco), orchestrating 4 specialized AI agents Planning, Market Intelligence, Opportunity Discovery, and Executive Briefing to take a founder from a single goal to a full execution roadmap. Generated complete, structured roadmaps in under 10–15 seconds end-to-end; validated across 3+ real startup use cases. Designed structured output validation between each agent layer.

Next.jsTypeScriptInsForgeMulti-Agent Orchestration
View all work

Wanderlust AI - AI Travel Planning Application

Built and deployed full-stack AI travel planner on GCP using Google Vertex AI (Gemini) with Supabase for user auth, database management, and persistent state. End-to-end API pipeline from user input to personalized AI-generated itineraries with real-time accommodation and car rental recommendations; live in production.

Vertex AI (Gemini)GCPSupabasePostgreSQL
View all work