I turn 'someone should build this' into 'oh, it's already done.'
Yashwanth Reddy Katipally - AI Engineer
I turn 'someone should build this' into 'oh, it's already done.'
I chase the hard problems no one wants. Then I ship the fix.
About
I'm an AI Engineer focused on building robust, end-to-end multi-agent systems.
I build custom agent loops, structured tool-calling surfaces, and MCP integrations that deploy across any surface - terminal, cloud, macOS, iOS, and Electron. Tech-agnostic and highly adaptable, with a strong focus on core problem-solving over framework loyalty.
What I do /
( SERVICES )
Agent systems don't happen by chance, they are built with intention. I engineer robust, scalable multi-agent solutions that make complex workflows effortless.
Agent Systems
I build custom-coded multi-agent systems tailored to your workflow. I focus on making them scalable, fast, observable, and equipped with engaging real-time interactions to provide a reliable experience for users.
01Custom Agent Loops
02Real-time Streaming
03Scalability
Advanced Solutions
I specialize in developing advanced AI applications that redefine the possibilities of online platforms, crafting local-first agent desktops, macOS integrations, and complex multi-provider routing systems.
01Local-First Apps
02Complex Integrations
03Innovative Tech Solutions
Code Optimization
Your agent system deserves to be fast, reliable, and cost-efficient. By optimizing inference times, improving context compaction, and routing fallbacks, I ensure your stack runs smoothly and efficiently.
01Modern Frameworks
02Context Compaction
03Speed Optimization
Things I've Built
A collection of agent systems, desktop products, and AI tooling shipped across terminal, cloud, and native surfaces.
The Path
My education and experience - the path that shaped my approach to building multi-agent AI systems.
2020
Started B.Tech in Computer Engineering with an AI/ML Specialization at Presidency University, Bangalore. Built the foundations of programming, data structures, algorithms, and machine learning.
2024
Completed B.Tech and began M.S. in Applied Data Intelligence at San Jose State University. Coursework in Generative AI, Machine Learning, Deep Learning, and Distributed Systems.
2025
Software Engineering Intern at Datasaur (San Francisco, CA). Shipped a conversational workforce AI agent (LangGraph, FastAPI, React) unifying Slack, Gmail, Notion, and Calendar - owned the tool registry and OAuth 2.0 scope design, cut median task time from 8 minutes to under 1. Built a hybrid pgvector + BM25 retrieval pipeline lifting recall@5 from 0.72 to 0.89. Deployed on AWS EC2/RDS with Docker and CI/CD, holding p95 inference under 1.2s and the daily token bill under five dollars.