Nithin Makkina

About Nithin Makkina

I am a Generative AI Full-Stack Engineer with 3+ years of experience building AI-driven systems across healthcare, pharma, and cloud environments. My work focuses on delivering production-grade LLM solutions—Chatbots, RAG pipelines, vector search engines, document intelligence workflows, and scalable backend services using Python, FastAPI, SQLAlchemy, and multi-cloud platforms.I work across the full AI lifecycle, from data processing and embeddings to retrieval frameworks, chatbots, PII-safe pipelines, SQL agents, and model monitoring. I’ve delivered multiple enterprise solutions for IBM that improved decision-making, accelerated insight generation, and automated complex manual processes.I bring a balance of strong engineering fundamentals and applied AI expertise via microservices. I take ownership of problems end-to-end, collaborate closely with cross-functional teams, and design systems that are reliable, efficient, and aligned with business strategy. My goal is to build AI products that create measurable impact at scale.

Experience

1. Gen Ai Full Stack Engineer Sde- (Via Randomtrees)

   IBM

   Oct 2023 — Present · US

   Deliver AI-driven platforms and LLM applications for enterprise teams across Pfizer, enabling faster decision-making, improved operational efficiency, and automation of previously manual, high-effort workflows. Lead end-to-end development across RAG pipelines, backend systems, document intelligence, and multi-cloud AI integrations.Customer Insights •Built a GenAI-powered insights chatbot using RAG, embeddings, Snowflake, and VOX. •Reduced insight generation time from weeks to minutes, accelerating CX decision-making, improving research efficiency by 60%. •Automated extraction, summarization, sentiment, and theme analysis over cross-country markets. •Improved data processing accuracy and consistency by 35%+ through PII-safe pipelines.Individual Case Safety Reports Automation – Prompting, Testing & Automation •Created prompts for high-accuracy medical entity extraction using Azure Prompt Flow. •Enhanced reliability of ICSR extraction workflows and reduced turnaround time for pharmacovigilance processing. •Automated JSON vs. Argus validation, increasing coverage and reliability by 70%. •Designed nested-loop algorithms for end-to-end ICSR testing automation.Theia – Document Intelligence Chatbot •Developed backend for a multi-source RAG chatbot using FAISS, Snowflake, Neo4j, and PostgreSQL. •Reduced manual document review time by 40–60%, improving grant and RFP workflows. •Delivered high-accuracy retrieval, summarization, and citation capabilities.EP & S – Incident Intelligence Platform •Redesigned and developed backend code, improving performance and reducing latency by 65%. •Enabled faster triaging of operational incidents, also improved analyst productivity and reduced manual investigation time. •Developed AI-driven incident extraction and classification to support operations teams. •Delivered secure, scalable APIs with SSO and multi-source ingestion.

Education

• University of Windsor

  Master's degree, ELECTRICAL AND COMPUTER ENGINEERING

• Velammal Global School

  High School Diploma, Computer Science

• SRM IST Chennai

  Bachelor of Technology - BTech, ELECTRONICS AND COMMUNICATION ENGINEERING