Akhil Karra
Class of 2028
Contact
- Email: akarra@andrew.cmu.edu
In This Section
- Academics
- Admissions
- Careers
-
News
- 2025 MSCF Trading Competition
- A New Academic Year at MSCF
- Alumni Reflect on the MSCF Program’s 30th Anniversary
- Breaking Barriers, Building Leaders: Women in Quant Finance
- Data Science in Finance
- Financial Engineering Salary
- High Stakes and Fair Values: CMU Students Face Off in the 2026 Market Making Game
- How to Become a Quant
- Introducing the MSCF Quantitative Assessment
- MSCF 30th Anniversary Celebration
- MSCF Advisory Board Member, Roni Israelov receives the 2024 Peter L. Bernstein Award
- MSCF Hosts 2025 Panel for Women in Data Science Pittsburgh
- MSCF Hosts 2nd Annual Datathon: Advancing Experiential Learning and Industry Connections
- MSCF Welcomes Rhonda Khan as Communication and Leadership Instructor and Coach
- MSCF Welcomes Shelli Faber as Associate Director of Career Services
- Quantbot Classroom Naming
- Squarepoint Foundation Deepens Partnership MSCF Through $100K Gift to Support Future Leaders
- Our Community
- Student Experience
Biography
I’ve been flying airliners in flight simulators since I was six, teaching myself new aircraft types from the manuals. Understanding a complex system enough to model how it behaves, and where it breaks, is what pulls me toward quantitative research. I’m a CMU MSCF student with a background in math, statistics, and AI. I interned twice at the Federal Reserve Board, where I built LLM tools still used for economic research, and I co-authored a published paper on where LLMs fall short at forecasting. That paper got me hooked on creating trustworthy research tools. Now I pair asset pricing and portfolio optimization with proofs in Lean 4, a proof assistant, so a model’s key properties hold for every input, not just a list of test cases. These proofs act as collateral behind a team’s trust. Building that trust in a team is vital to me. I learned how I like to work in high school music: conducting at Carnegie Hall, I workshopped the score with the musicians rather than dictating it. This is still how I picture a research team: building something none of us could reach alone. If that’s the kind of team you’re building, I’d love to talk.