Anant Tyagi

About Anant Tyagi

I am an AI/ML Engineer with over 4 years of experience designing, developing, and…

Experience

1. Ai Ml Developer

   Chirpn It Solutions

   Jan 2025 — Present · Fremont, CA, US

   Developed hybrid recommendation engine with ALS and BERT embeddings, increasing click-through rate by 17% and order value by 12%. • Built LLM-powered workflows using OpenAI GPT-4, LangChain to auto-extract structured data from documents, boosting accuracy. • Implemented PySpark pipelines with Great Expectations to validate and process clickstream and profile datasets, ensuring high-quality inputs for personalization models. • Deployed FastAPI-based recommendation and LLM APIs on Kubernetes, sustaining p95 latency under 100ms during peak traffic events. • Designed Redis-backed feature store with Kafka streaming, enabling real-time, session-aware personalization across millions of users. • Developed CNN-based defect detection model with PyTorch and Optuna, achieving 90% accuracy in production testing. • Automated GenAI cost tracking with Python and Jinja2, reducing reporting and analysis time by 40%. • Established MLflow tracking and centralized model registry, ensuring reproducible workflows and safe blue-green deployments. • Conducted controlled A/B testing with sequential monitoring to validate significance and optimize recommendation performance. • Integrated bias mitigation algorithms into ranking pipelines, ensuring fair exposure across demographics and reducing skew.

Education

• The University of Texas at Dallas

  Master's degree, Computer/Information Technology Administration and Management

• GITAM Deemed University

  Bachelor of Technology - BTech, Engineering