Senior Software Engineer Β· AI Engineer β Toronto, Ontario, Canada π¨π¦
I build software end to end: AI-powered products, agentic workflows, and full-stack applications β from prototype to production. 6+ years across backend, frontend, and cloud infrastructure, currently focused on LLM applications and agentic AI.
π€ payment-assistant β I taught Stripe to understand English
A full-stack AI payments operations assistant: business owners manage payments, refunds, and invoices in plain language β "refund Maya Johnson's last payment, then send me today's summary" becomes validated Stripe API calls. The LLM interprets intent, deterministic Python validates and executes, and a $2,000 safety rule keeps the AI from getting generous with your money. Customers get their own privacy-preserving Telegram bot that only ever shows them their own data.
Python FastAPI React TypeScript OpenAI Stripe Telegram
π network-speed β speed tests for people who fear the terminal
A Windows desktop app in Rust that wraps iperf3: configure a test, watch live throughput draw itself on a chart, cancel mid-run, and export results to CSV/JSON/TXT. Auto-detects iperf3 (or bring your own binary), handles every failure mode with a human-readable message, and ships as a one-click NSIS installer with iperf3 bundled β no terminal, no setup.
Rust egui NSIS
π React-table β a data grid that doesn't choke
A React 18 + TypeScript table on TanStack Table v8 with virtualized scrolling that stays smooth past 500 rows, sortable computed columns, drag-and-drop header reorder, and localStorage persistence.
React TypeScript TanStack Table Vite
π takehome β JSON in, answers out
A Ruby script that crunches user and company datasets into token top-ups and email eligibility β paranoid about bad data, deterministic in its sorting, and verified by diffing against expected output.
Ruby
π codecraft-masterclass β learn it, then ship it
Production-ready source code, architecture diagrams, and runnable examples for CodeCraft Masterclass tutorials: LangGraph multi-agent systems, production RAG pipelines, full-stack AI with Next.js 15, Go microservices & concurrency, and distributed system design.
Python LangGraph Next.js Go
Languages: Python Β· TypeScript Β· Ruby Β· JavaScript Β· Rust Β· Go Backend: FastAPI Β· Node.js Β· Ruby on Rails Β· PostgreSQL Β· Redis Β· MongoDB Frontend: React Β· Angular Β· Vite AI: LLM applications Β· Agentic workflows Β· RAG Β· Vector databases Β· Prompt engineering Cloud & DevOps: AWS Β· GCP Β· Docker Β· Kubernetes Β· CI/CD
- πΌ LinkedIn
- βοΈ komirishettyajay1@gmail.com
Build something simple first. Measure what doesn't work. Iterate quickly. Then make it scalable.
