About

AI Engineer with 2+ years of experience owning the R&D, planning, architecture, and end-to-end implementation of production-grade Generative AI applications for multiple enterprise clients. Specializes in LLM-powered applications, RAG systems, agentic workflows, multimodal AI, and Voice AI, integrated with scalable backend services and business platforms.

Experience1

  1. Full Stack Generative AI Engineer

    Jan 2024 — Present

    Noesis.tech

    Engineered an enterprise RAG pipeline utilizing LlamaParse, Weaviate, and Gemini to ground LLM responses in proprietary data, cutting agent hallucinations by 75–85%. Optimized AI inference speed by up to 3x with a Redis-backed state management system. Designed agentic AI workflows and architected a multi-LLM gateway to support various AI models, enhancing customer interactions across multiple platforms.

Projects5

  • Enterprise RAG & Semantic Search Platform

    Engineered a two-stage Weaviate retrieval pipeline improving Recall@10 by 14% and re-architected ingestion using LlamaParse to lift LLM accuracy from 64% to 89%. Integrated Ragas into CI/CD to systematically track Context Precision, reducing production hallucinations by 80%.

    • Weaviate
    • LlamaParse
  • Real-Time Voice AI Microservice

    Engineered a bidirectional Voice AI pipeline for autonomous calling, reducing manual sales agent workload by 70–75%. Implemented low-latency dialogue mechanics for enhanced user interaction.

    • Voice AI
    • Plivo
    • Azure Speech
  • Autonomous AI Sales Agent

    Engineered an automated AI workflow across WhatsApp and Instagram to process inquiries, slashing manual workload by 85–90% and delivering instant engagement to accelerate conversion cycles.

    • WhatsApp
    • Instagram
    • Shopify GraphQL
  • Email Agent Integration & HITL

    Engineered a scalable email integration utilizing OAuth 2.0 for automated sales resolution, significantly optimizing LLM inference and API polling costs.

    • OAuth 2.0
    • Gmail
    • Outlook
  • Enterprise AI Credit Billing & Metering System

    Architected a universal credit-based billing engine to normalize usage metrics, ensuring predictable margins and eliminating conversational latency bottlenecks.

    • MongoDB
    • Redis

Education1

  • Bachelor of Technology (B.Tech)

    2020 — 2024

    KJ Somaiya Institute of Technology

    Artificial Intelligence and Data Science