Praveen Kumar

Principal AI Engineer & Solution Architect

I build end-to-end AI products, from first prototype to production. I'm deep into speech AI, OCR, and LLMs, and I love sharing what I learn through open-source projects and blogs.

View My Work

About Me

I'm an AI Engineer focused on building systems that actually work in the real world. My background is in speech intelligence, LLM automation, and document processing. I don't just build models; I build the infrastructure that makes them reliable and fast.

I've spent the last few years architecting AI platforms for enterprise clients, from real-time voice agents to massive document intelligence suites. I like taking messy, complex ideas and turning them into clean, production-ready code that teams can actually trust.

When I'm not coding, I'm usually mentoring teams or figuring out how to make GPU workloads more efficient. I'm big on modular design, observability, and building solutions that align with real business goals.

Python PyTorch FastAPI LLMs & RAG Speech AI vLLM Whisper LangChain OCR & Document AI Vector Databases Docker & K8s AWS & GPU Infra PostgreSQL Redis Multi-Tenant SaaS CI/CD
10+
Production AI Platforms
5+
Years Experience
50+
Models Deployed
15+
Enterprise Clients

Featured Projects

Context Search Engine

Context Search Engine

AI-powered semantic document search platform using transformer embeddings and FAISS. Goes beyond keywords to understand meaning and context in your documents.

Transformers FAISS NLP
Docuvera - Loss360

Docuvera - Loss360

Enterprise multi-tenant SaaS platform for automated insurance loss run processing. Combines OCR and domain-specific LLMs to extract structured data from unstructured documents.

OCR LLM Multi-Tenant InsurTech
Movie Recommendation System

Movie Recommendation System

Production-ready recommendation engine using TF-IDF, SVD, and content-based filtering. Scales from 10K to 930K+ movies with intelligent quality filtering and Django deployment.

ML TF-IDF SVD Django

Latest Blog Posts

🔍
December 2, 2025

Building a Hands-On Semantic Search Engine

From scratch to search: A complete walkthrough of building a semantic document search engine with embeddings, FAISS, and transformers...

Read More →
🏗️
December 2, 2025

Building Loss360: Architecting a Multi-Tenant AI Platform

Deep dive into architectural decisions, technical challenges, and lessons learned from building an enterprise AI document processing platform...

Read More →
🎬
December 5, 2025

Building a Scalable Movie Recommendation System

Deep dive into building a production-ready recommendation system using TF-IDF, SVD, and content-based filtering. Scale from 10K to 1M+ movies...

Read More →

Let's Connect

Have a cool project in mind or just want to chat about the latest in AI? I'm always up for a good conversation. Drop me a line!