I Build AI
Systems That
Actually Ship.
I build RAG systems, AI agents, and MLOps pipelines that actually make it to production — typically in under 4 weeks.
About
Krish
My Approach
I treat every ML system like infrastructure, not a science experiment. That means drift detection, failure fallbacks, cost guardrails, and clean documentation — so your team owns it long after our engagement ends.
Where I Come From
Mathematics and software engineering background. I've shipped computer vision pipelines, on-device OCR apps, multi-agent systems, and enterprise RAG architectures — both solo and embedded with product teams.
What You Actually Get
Weekly demos, clear communication, and no disappearing after deployment. You get a documented, monitored, maintainable system — with observability baked in from day one. Reliability isn't a premium add-on. It's the baseline.
Professional
Experience
Freelance ML & Software Engineer
Ship production-ready ML systems for startups and engineering teams — RAG pipelines, autonomous agents, and MLOps infrastructure. Every project includes monitoring, documentation, and full handover.
Machine Learning Engineer Intern
Built production-ready salary prediction and fraud detection models using scikit-learn pipelines — from feature engineering through evaluation and deployment.
Flutter Developer Intern
Built the ML integration layer for SnapText, an on-device OCR app — enabling real-time text extraction without cloud dependency.
PHP Developer Intern
Built full-stack web modules using PHP, MySQL, and JavaScript — foundational engineering experience that shaped my approach to production systems.
Core
Services
Enterprise RAG
Your team shouldn't spend hours hunting through scattered documents and getting wrong answers. I build RAG systems with hybrid search, reranking, and evaluation built in — so the right answer surfaces in seconds, not hours.
- Self-Querying & Hybrid Search
- Advanced Chunking Strategies
- RAGAS Evaluation & Observability
- Multi-Vector Indexing
Agentic AI
From internal research assistants to customer-facing automation — I build multi-agent systems that handle complex workflows while keeping humans in the loop where it matters. Monitored, production-deployed, with proper safeguards.
- LangGraph/CrewAI Frameworks
- Function Calling & Tool Use
- Memory & State Management
- HIL (Human-in-the-loop) Design
MLOps Architect
Models that only work in notebooks don't generate revenue. I build the full ML lifecycle — reproducible pipelines, experiment tracking, drift detection, and CI/CD for ML — so your models stay accurate long after go-live.
- CI/CD for Machine Learning
- ZenML/MLflow Infrastructure
- Model Monitoring & Drift Detection
- Kubernetes/Cloud Orchestration
Selected
Case Studies
AmberClaw — AI Memory Assistant
Salary Predictions For Data Professions
SnapText — OCR App
Fitness Tracker ML Analysis
EchoNest — Social Media
All-In-One Discord Bot
Technical
Resources
Certificates
Industry-recognized certifications in AI, ML, NLP and more.
MLOps
Machine Learning operations, DSA, NLP, and Neural Networks resources.
Cloud
Cloud computing, security, virtualization, and data center guides.
CyberSecurity
Cryptography, ethical hacking, network security, and forensics.
DBMS
Database management — SQL, normalization, ER diagrams, and transactions.
DevOps
Software development lifecycle, Agile, SRS, and risk management.
BlockChain
Introduction to blockchain technology and distributed ledger concepts.
Digital Marketing
Digital marketing strategy, entrepreneurship, and GPT prompting guides.
Trusted by
Industry Leads
"Krish delivered a production RAG system in under 3 weeks. The handover documentation was thorough enough that our junior engineer could maintain it independently."
"His MLOps setup saved us from what would have been weeks of manual pipeline work. Everything was tracked, versioned, and reproducible from day one."
Let's Build
Something
That Works
Got a project in mind? Tell me what you're working on and I'll get back within 24 hours with honest thoughts on fit, scope, and timeline.