Hire Prithviraj P. - LLM Engineer

Designed and developed a browser-based 3D CAD viewer capable of visualizing industry-standard engineering models entirely client-side, eliminating the need for server-side geometry processing or proprietary plugins. Integrated the geometry kernel, enabling STEP file parsing with full assembly tree extraction, part metadata (names, instance hierarchy, colors), and support for complex engineering model structures. Architected a high-performance rendering pipeline using WebGL, implementing orthographic viewing modes and an interactive navigation system similar to professional CAD viewers. Optimized application performance by executing computationally intensive geometry processing within Web Workers, allowing complex assemblies (70+ components) to be parsed and rendered smoothly at interactive frame rates (~60 FPS). Implemented structured extraction of product design data from CAD assemblies, enabling downstream integration with AI systems for design intelligence, part classification, and automated engineering workflow generation. Demonstrated how complex engineering data and computational workloads can be processed directly in modern browser environments using WebAssembly-based architectures. Tech Stack: React, TypeScript, Babylon.js, WebGL, Vite Architected and delivered a multi-tenant AI SaaS platform for document intelligence, reducing manual search and analysis time from hours to seconds (~70–90% time savings). Designed a secure, scalable cloud architecture supporting concurrent users with strict data isolation and role-based access control, enabling enterprise-grade reliability and compliance. Improved response consistency by implementing conversation memory and prompt versioning, enabling continuous optimization of AI outputs in production. Optimized system performance and reduced LLM costs by ~20–30% through efficient retrieval and prompt strategies. Tech Stack: OpenAI, Llama, LangChain, FastAPI, Next.js, PyPdf, OCR, AWS, Pinecone, PostgreSQL

* Designed and delivered AI solutions for a U.S. Employee Benefits SaaS platform (SAP BTP), improving access to HR and benefits information and reducing manual effort for HR teams and employees. Built a natural language-to-DSL engine (Elasticsearch query generation), enabling non-technical users to create complex queries from simple English — reducing query creation time from minutes to seconds and minimizing reliance on technical teams. Developed a Helpdesk AI agent using advanced RAG, enabling secure conversational access to private HR and benefits data — reducing support resolution time by ~40–60% and improving employee self-service. Designed an AI-powered data mapping solution for EDI 834, automating schema mapping between insurers and employers — reducing manual mapping effort by ~50–70% and significantly lowering data errors. Built AI chatbots for HR policies and benefits documentation, reducing repetitive HR support queries by ~30–50% and improving response consistency. Developed a web/URL-based AI assistant for IRS-related queries, enabling faster access to external regulatory information and reducing time spent searching across multiple sources. Delivered multiple AI solutions that improved operational efficiency, reduced manual workload, and enhanced employee experience across HR and benefits workflows. Tech Stack: OpenAI, LangChain, Node.js, PyMuPdf, OCR, Elasticsearch, SAP BTP

* 𝗚𝗲𝗻 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁 𝗗𝗲𝘃𝗲𝗹𝗼𝗽𝗺𝗲𝗻𝘁 * HR Customer support AI Agent: Auto email response generation and auto ticketing using LangGraph. * Email AI Chatbot: Intelligent search system to retrieve decisions and data from 21,000+ email threads. * Fashion ShopAssist: Personalized fashion recommendation chatbot using a catalog of 14,000+ products. * Insurance AI: Generative AI for policy queries, handling complex PDFs in conversational style. * Invoice Analysis AI: Chatbot for invoice-based Q&A using OpenAI, Gemini API, and NLP. * Laptop Recommendation Bot: AI-driven chatbot recommending laptops tailored to user needs. * Mobile Apps: App for students to practice mock tests on the go. App for personal money management and mobile banking. App for connecting and using microcontroller boards like Arduino on bluetooth.