Aniket Poojari

About Aniket Poojari

AI/ML Engineer with 3+ years of experience building Agentic AI systems, LLM applications, and scalable MLOps pipelines. Focused on shipping production-ready AI that is fast, reliable, and measurable.What I built- Agentic AI systems with LangGraph, MCP, and multi-tool orchestration- LLM applications: RAG chatbots, fine-tuning pipelines (LoRA/PEFT), guardrails, red-teaming- Distributed training pipelines on AWS SageMaker with DeepSpeed- End-to-end MLOps with DVC, MLflow, CI/CD, and automated quality gatesImpact delivered- 66% faster support resolution via on-premise RAG chatbot at Kritova- 3x training speedup using DeepSpeed distributed training on SageMaker- 25% LLM cost reduction through token analytics and model caching- 94.2% code generation pass rate via dual LoRA fine-tuning on Phi-3- 99.5% adversarial inputs blocked via multi-layer guardrails (prompt injection, PII, SQL injection)- 216 adversarial attacks across 6 categories in a full-stack LLM red-teaming platformCurrently building demand forecasting and supply chain ML systems at Southern Wholesale of Florida.Open to AI/ML Engineering roles. Let\'s connect: a••••••••@gmail.com

Experience

1. Machine Learning Engineer

   Kritova

   Mar 2025 — Present · US

   Reduced internal support resolution time by 66%(3 days to 1 day) by deploying a secure on-premise RAG chatbot while maintaining 100% data privacy compliance- Built an end-to-end ML-integrated platform with 4 microservices serving 300+ concurrent users, implementing automated data pipelines for course analytics and real-time performance tracking across 50+ learning modules.

Education

• Fr. Conceicao Rodrigues College of Engineering

  Bachelor of Engineering - B.E., Information Technology

• Oregon State University

  Master of Science - MS, Computer Science