Ananth Sriram
I am a CS + ML student at UMD (Class of 2026). My work focuses on multi-armed bandits and remote sensing, with current research in Columbia's CRIS lab and UMD's Digital and Precision Agriculture Lab. Previous industry experience at Capital One and ZoomInfo. Always looking to expand research and industry experience, contact me!
News
- May ’26: Released our IronSite safety assistant architecture preprint on arXiv.
- Oct ’25: Accepted Full-Time offer from Capital One.
- Sep ’25: Started as Research Assistant at Columbia University CRIS.
- Jun ’25: Joined Capital One as a Software Engineer Intern (Finance Tech).
- Jul ’24: Published “An Overview of Drones in Agriculture” in UMD Extension.
Experience
Research
- Columbia University (CRIS) — Research Assistant. Explainable AI (XAI) for complex scientific and physicochemical systems.
- University of Maryland (DPAL) — Research on precision agriculture and remote sensing; co-authored extension publication on drones in agriculture.
- University of Maryland (TEBL) — Undergraduate Research Assistant (Jan 2024 – May 2024). Integrated a predictive ML model into a real-time analytics platform to improve the reliability of a physical bioprinting system.
- University of Maryland — Department of Geographical Sciences — Undergraduate Research Assistant (Sep 2023 – Dec 2023). Benchmarked CNN/SVM approaches for UAV-based fire detection (XPRIZE challenge).
- Climate Computing (FIRE) — Undergraduate Research Assistant (Aug 2022 – Nov 2023). Processed and analyzed terabyte-scale datasets in a supercomputing environment using Python and shell scripting.
Industry
- Capital One — Software Engineer Intern, Finance Tech (Summer 2025).
- ZoomInfo — Software Engineering Intern, MLOps (Summer 2024).
Publications & Projects
- Passive Construction Site Safety Monitoring via Persona-Scaffolded Adversarial Chain-of-Thought VLM Verification (Preprint, May 2026). First place at the UMD Spatial Intelligence Hackathon. arXiv / Code
- An Overview of Drones in Agriculture (UMD Extension, Jul 2024). Paper / resource
- AI Music Detection — Full-stack app: spectrogram CNN (92% accuracy), React frontend, Flask backend. Code
Out and about
A few places off-screen. Check out my rate your music! Rate Your Music
