Stanford Code in Place
Led a 10-15 student remote section for Stanford Code in Place, walking Python from problem statement to readable code. Main focus on making decisions under uncertainty and treating bugs as information, not just perfect syntax.
HARSH
KUMAR
Building production RAG systems that ship reliably, fail gracefully, and behave predictably under real load.
Hi, I'm Harsh. I'm a software engineer. I write backend code and build features on top of AI models.
I got into computers as a kid, fixing up an old PC so it could run games. I've been tinkering ever since.
A few things I've worked on:
- a zero-downtime database migration
- prompts running on six OpenAI models
- REST APIs on AWS Lambda
I'm looking for my next role. If any of that sounds interesting, say hello.
I teach
Still learning, sometimes from the front of the room. These days that means a small Code in Place section on Zoom; earlier, it meant classrooms at ASU where the goal was to make web dev, algorithms, Git, and beginner CS less intimidating.
Software Developers Association (SoDA), ASU
As Technical Development Director for SoDA, I led a team of 5 to design and run weekly workshops for 90-100 ASU students - competitive programming, web dev, interview prep, tooling, beginner CS - and ran live coding sessions for our Leetcode meetings. I also spearheaded the 2022 Annual Coding Bootcamp, a 2-hour hands-on seminar in Kotlin, XML, and Android Studio for 150 participants, where 92% successfully built and deployed their first Android app."
* When not debugging, probably listening to:
Recently played
Numb
Marshmello, Khalid
** Or silence. Debugging requires silence.