Sarvesh Ss

Designed agentic AI systems to automate internal workflows, achieving a measured 38% reduction in wall-clock time across three workflows. Containerized ML inference services on AWS and fine-tuned classification models in PyTorch with automated monthly retraining via S3 event triggers.

Built a RAG application over a 50K-document corpus and load-tested performance to confirm sub-300ms P95 latency at 20 concurrent users. Delivered an Egkam Spices & Herbs e-commerce site with reported 3× increase in online orders within 60 days.

Built Azure AI Foundry models for document classification and sentiment analysis, reducing a client’s monthly manual review workload for routine categories. Developed Power Automate/Dataverse automations and Copilot Studio agents to resolve Tier-1 support queries and reduce escalations.

Built ID/admit card generation modules in PHP/MySQL for the GraceER ERP project, reducing admin processing time per cycle. Developed an autonomous campus robot (MobileNetV3 facial recognition, Whisper voice interaction, GPS navigation) and deployed an end-to-end RAG system using Ollama (Llama-3) with MongoDB Atlas Vector Search.

Grade: GPA 9.38/10

Activities and societies: Top 5% academic standing in the B. Tech cohort; consistently on Anna University merit list.

Pursuing a B. Tech in AI & Data Science at Anna University since Nov 2022. Earned a GPA of 9.38/10 and ranked in the top 5% of the 2024 cohort.