Archit Gupta went from never having written a line of code to building AI tools that grade hundreds of student cases and a Socratic AI tutor for a graduate class — in one year.
Archit is graduated from Foster’s MSBA (Master of Science in Business Analytics) in June, and served on the committee that planned Foster’s first AI Spark Day. Before Foster, he worked in M&A at KPMG, founded a consumer fintech startup focused on retail investors, and held product roles at a Singapore-based fintech. He entered the program with no software background and built AI-assisted grading and tutoring tools at UW with Professor Leonard Boussioux. He also founded a student club to pass institutional knowledge between Foster’s one-year specialty master’s cohorts.
- How to get past the AI “cold start” when you’ve never touched a command prompt
- How to use AI for grading at scale while staying FERPA-compliant (anonymization + open-source models + human in the loop)
- Why a Socratic AI tutor can protect learning where an unrestricted LLM undermines it
- How to design an A/B test that measures retention, not just scores
- Why “everything that can be automated need not be automated” — and how to tell the difference
- What rapid iteration looks like when you don’t need an engineering team
Resources mentioned: n8n; Claude Code; open-source / frontier LLMs; FERPA; Ryan Holiday’s newsletter; Foster’s AI Spark Day.