Kirtan Thummar
AI Orchestration & Cloud Engineering
I'm Kirtan — an engineer who builds developer tools, AI workflows, and cloud systems. Right now I'm working on an autonomous agent orchestrator for software development, an AI-powered GitHub code reviewer at ASU's LEI lab, and a handful of cloud automation projects. Open to roles where I can keep building useful things.

A Bit About Me
I think AI is a baby with superpowers. It doesn't know how to use them yet, and it'll do exactly what its guardians teach it to do. We're collectively expecting too much from it without really understanding what it's actually good at — and that gap is where most of the interesting problems live.
That's a big part of why I started Vctrx. The problem I keep running into, across every AI project I've worked on, is memory and context. AI that forgets where it is, who it's talking to, what happened three steps ago. I want to fix that for software teams — starting from the actual problem and iterating all the way to real user feedback.
Right now I'm deep in building an AI orchestration harness using tools like LangGraph. The idea is a structured layer that manages how agents hand off work, retain context across long tasks, and stay on track through an entire development workflow — not just a single prompt-response loop. It's the thing I wish existed when I started building with AI, so I'm building it.
I'd rather work with a team that believes quality and speed can coexist than one moving fast purely out of FOMO.
Tech Stack & Expertise
Technologies I work with daily


What I Do
My areas of expertise
AI Agent Orchestration
Building structured layers that give AI agents memory, context, and the ability to hand off work across long multi-step workflows — using tools like LangGraph.
AI Orchestration & Harness Development
Designing multi-agent systems that retain context, coordinate handoffs, and stay on track across complex workflows. The glue layer between raw LLM capability and real-world software tasks.
LLM-Powered Developer Tooling
Turning AI into practical dev tools — automated code reviewers, prompt pipelines, evaluation loops. Things that slot into existing workflows rather than replacing them.
Cloud Infrastructure & AWS
Serverless APIs, event-driven pipelines, and IaC with Terraform and CDK. Mostly AWS — Lambda, ECS, DynamoDB — with a focus on keeping things simple and cost-efficient.
CI/CD & DevOps Automation
GitHub Actions pipelines, automated deployments, environment management. Making the path from code to production boring in the good way.
Platform Engineering
Internal tooling and developer platforms — infra that other engineers actually want to use, not fight against.
Let's Connect
Have a project in mind? Let's collaborate and create something amazing together
kthumma5@asu.edu
Location
Tempe, AZ • [ Open to relocate ]