Senior Software Engineer turning caffeine into AI-powered solutions at Sanas AI

I build things that think. From RAG-based chatbots helping law enforcement query evidence to real-time LLM engines that make sales teams unstoppable. If it involves AI, scalable architecture, or making computers do impossible things, I'm probably knee-deep in it.

Right Now: Architecting unified AI platforms that replace janky CLI tools, optimizing Triton Server inference pipelines, and building LLM-powered productivity tools that actually work in production.

The Stack I Live In:

• AI/ML: LLM, RAG, Langchain, LangGraph, OpenAI API, Vector DBs (Qdrant, Weaviate, Neo4j), Triton Server
• Backend: Python (FastAPI, Django), Node.js (NestJS), Golang, PostgreSQL, Redis
• Frontend: React, Next.js, TypeScript, TailwindCSS, Storybook
• DevOps: Kubernetes, Docker, AWS, Azure, CI/CD pipelines

🚀 Unified AI Platform @ Sanas
Replaced fragmented tools with a centralized hub. Real-time job tracking, automated Colocation-to-S3 pipelines, batch inferencing on Triton for ultra-low latency demos.

🤖 RAG Systems for the Real World

• Built conversational chatbot for ITER: massive datasets, high accuracy, automated citations
• Created universal file processor: any media → any vector DB
• Shipped semantic search for NYPD: natural language queries on video/audio evidence

📐 Babson College Mathbot
RAG-based agent with custom LaTeX renderer. Students ask math questions, get visual solutions. Backend's so modular you can plug it into any vector DB.

⚡ AI Productivity Tools

• Sales enablement with live transcription + LLM suggestion engine
• InfoSec Slackbot automating doc Q&A with OpenAI
• Gratitude bot generating personalized avatars via Google Imagen

• 60% productivity boost for Customer Success in 3 months
• 80% increase in customer acquisition (POC → production in 6 months)
• 2 months to ship production SaaS with 4-person team
• Quarterly Team Award + multiple Monthly Performance Awards

Watched Iron Man as a kid. Wanted to fly. Ended up making computers smarter instead.

Started with Windows 98, dabbled in assembly, built online compilers in college. Evolved through MERN stack, DevOps, and cloud platforms. Then AI happened and everything clicked. Now I'm building systems that help everyone from college students to law enforcement agencies.

The through-line? Endless curiosity for how things work and making impossible things possible.

Looking for roles where I can:

• Architect large-scale AI/ML systems
• Build developer platforms and internal tools
• Push LLM/RAG architectures to production
• Solve hard problems with smart code

Hit me up: bhaweshverma50@gmail.com