Hi, I'm Mani
const mani = AI Engineer @ Microsoft · Fullstack · Open Source
I don't just write code.
I build systems that matter.
I'm an AI Engineer with 9+ years building production-grade systems at Microsoft, Stripe, and Palantir. My work sits at the intersection of AI/ML engineering, fullstack development, and platform architecture — shipping real products that process millions of requests and generate real revenue.
AI Systems Architecture
Designing production LLM pipelines, RAG systems, and multi-agent frameworks that scale to millions of requests.
Fullstack Engineering
End-to-end product delivery from React frontends to FastAPI backends with clean architecture and strong testing practices.
Systems Thinking
Holistic view of software systems: performance, reliability, observability, security, and cost optimization at scale.
Rapid Execution
From whiteboard to production in days, not months. Shipping real systems with measurable business impact.
“The best AI system is the one that ships, serves users, and gets better over time. Everything else is theory.
”
Things I've Actually Built
Production systems, not side projects. Each one is live, battle-tested, and generating real value.
AI Document Intelligence Platform
Enterprise-grade document processing pipeline using Azure OpenAI, Form Recognizer, and vector embeddings for intelligent extraction and Q&A over 10M+ documents.
10M+ docs processed · <1s latency · 94% accuracy
Custom Vector Search Engine
High-performance semantic search engine built from scratch with approximate nearest neighbor search, HNSW indexing, and support for multi-modal embeddings.
10M vectors · <10ms p99 latency · 99.5% recall
My Technical Stack
Years of production experience across the full AI and engineering stack.
AI / ML
8 technologies
Frontend
8 technologies
Backend
8 technologies
DevOps
8 technologies
// skill levels based on production usage, not courses
Where I've Made Impact
Built and shipped production systems at some of the most demanding engineering organizations in the world.
Senior AI Engineer
Building next-generation AI-powered developer tools and enterprise solutions on Azure. Leading architecture design for large-scale LLM deployments serving millions of users.
- Architected a multi-tenant RAG platform processing 10M+ documents with 94% accuracy, reducing manual review time by 70%
- Led migration of legacy document processing pipeline to Azure AI services, cutting infrastructure costs by 45%
- Built an LLM evaluation framework used by 20+ internal teams to benchmark model performance across 500+ test cases
- Mentored 5 junior engineers and established AI engineering best practices adopted org-wide
Fullstack Engineer – AI Platform
Part of the AI Platform team building internal ML tooling, fraud detection systems, and developer-facing AI features for Stripe's global payments infrastructure.
- Shipped Stripe's AI-powered chargeback prediction model, reducing fraudulent disputes by 32% ($8M+ annual savings)
- Built a real-time data pipeline processing 50K+ transactions/second with ML scoring at <50ms p99 latency
- Developed the internal ML experimentation platform used by 40+ data scientists to A/B test models in production
- Contributed to Stripe's public Radar ML documentation and led 3 internal workshops on production ML systems
Software Engineer – Data Platforms
Worked on Palantir Foundry's data integration and transformation platform, building pipelines for government and enterprise clients processing sensitive, high-value datasets.
- Built a distributed ETL pipeline for a federal client processing 500GB+ of daily health data with 99.99% reliability
- Implemented role-based access control (RBAC) system securing data for 10,000+ users across 50+ organizations
- Reduced pipeline execution time by 60% through parallel processing optimization and intelligent caching
- Designed and shipped a self-serve data transformation UI adopted by 200+ non-technical users
AI Engineering Consultant
Provided AI engineering consulting services to early-stage startups, helping them design and implement ML systems, data pipelines, and AI-powered product features.
- Delivered 6 production ML systems for 4 startup clients across fintech, healthtech, and e-commerce verticals
- Built a NLP-based customer support automation system reducing ticket volume by 45% for a SaaS client
- Designed data architectures that scaled from 0 to 1M users without major rewrites
- Established engineering processes (code review, testing, CI/CD) for 2 early-stage teams
Building in Public
My open source work on GitHub. Real code, real commits, real impact.
github.com/maniscodebaseLet's Build Something Real
Have a project that needs serious AI engineering? I'm selectively available for the right opportunities.
Direct Contact
Typically responds within 24h
Whether it's a full-time role, consulting engagement, or collaboration on an interesting AI problem — I want to hear about it.