I'm the person who watched $47,000 disappear into an AI agent loop in production — and then built a company to make sure it never happens to anyone else.

Currently building GetOnStack — observability, loop detection, and cost control for multi-agent AI systems. Think Datadog, but for your AI agents that go rogue at 3am.

By day: Founding Engineer + grad student (somehow maintaining a 4.0 while doing both)
By night: Writing about the parts of AI nobody wants to admit out loud — 300,000+ reads and counting

I'm open to every role that lets me build real things with real impact — SWE, Platform Eng, AI Infrastructure, DevOps, full-stack, whatever. If you're building something interesting, I want to talk.

The company I'm building. Production observability for multi-agent systems.

What it does that nothing else does: detects agent loops at the conversation level before they drain your budget, attributes costs across every agent in a pipeline, and gives you the monitoring layer that LangChain and CrewAI ship without.

Born from a real $47,000 mistake. Built so you don't have to make yours.

Python GCP LangChain CrewAI Distributed Systems Observability

Kafka-inspired, built from scratch. Because sometimes you learn more by building the thing than reading about it.

Multi-broker replication, leader election, priority queues, consumer groups, real-time monitoring. Handles 100K+ messages/sec under load.

Go Distributed Systems Fault Tolerance Replication

Cloud-native multi-tenant SASE control plane. Zero-Trust policy engine + distributed edge gateways + real-time access enforcement.

Sub-10ms policy decisions via Redis. Full data isolation per tenant. The kind of thing that gets you past the HPE Juniper phone screen.

TypeScript Node.js Redis MySQL MongoDB

Enterprise-grade automated testing for network infrastructure. OSPF, BGP, MPLS validation with CI/CD integration. Built to speak network engineer fluently.

Python Pytest Network Engineering CI/CD

A production-grade legal document intelligence engine that ingests court opinions, extracts structured intelligence using LLMs, builds a citation graph to detect circuit splits, and exposes it all via a FastAPI service.

Python LangChain OpenAI Pydantic OCR

I write about AI infrastructure, production disasters, and the economics of building with LLMs. Real war stories. No fluff.

Published in Towards AI · Level Up Coding · Data Science Collective · Artificial Intelligence publication

📖 Read everything → medium.com/@tejakusireddy

AI/Agents    →  LangChain · CrewAI · OpenAI · Anthropic · RAG · Agent Observability
Languages    →  Python · Go · TypeScript · SQL
Cloud        →  GCP · Docker · Kubernetes · Redis · PostgreSQL · MySQL · MongoDB
Monitoring   →  Datadog · Distributed Tracing · Custom Observability Tooling
DevOps       →  GitHub Actions · CI/CD · Typed codebases · Test-driven everything

• Founding Engineer @ GetOnStack — building in production every day
• MS Computer Science @ Pace University, New York — 4.0 GPA (yes, while doing the above)
• Coursework: Distributed Systems · Computational Statistics · AI/ML · Cybersecurity
• Writing — next deep-dive incoming: architecture diagrams, real benchmarks, actual systems thinking

Open to any role where I get to build things that matter — SWE, AI/ML Engineering, Platform, Infrastructure, DevOps, full-stack, whatever. If the tech is interesting and the team ships, I'm interested.