Ashutosh Sundresh
Computer Science student at UCLA, open-source developer, and Kleiner Perkins Fellow. Building at the intersection of systems programming, mobile OS, and developer tooling.
About
Ashutosh Sundresh is a BS Computer Science student at the University of California, Los Angeles (2024–2028).
- Email: ashutoshsun@ucla.edu
- GitHub: https://github.com/AshutoshSundresh
- LinkedIn: https://linkedin.com/in/asund
- Twitter/X: https://x.com/asundresh
- Website: https://ashutoshsundresh.com
Website Map
- Home: Overview, current work, and contact links.
- Experience & Projects: Work history, projects, education, awards, publications, and activities.
- Coursework: UCLA course history and academic planning.
Work Experience
Software Engineering Intern at Brain Co.
- Location: San Francisco, California, United States
- Period: Jun 2026 - Sep 2026
- Link: https://brain.co
- AI Platform Team, joined as part of the first ~40 engineers, cross-cluster LLM inference metering platform + sovereign UAE region buildout
- Built cross-cluster, multi-provider LLM inference metering platform governing $100K+/mo spend across 170+ employees, 500+ API keys, and 4 Kubernetes clusters, using cross-subscription Azure Private Link and ArgoCD GitOps deploys
- Designed per-model, per-token-type allocator with cent-preserving math reconciling per-key cost to provider invoices
- Shipped an org-wide usage MCP server fronting the org's Google login via WorkOS, reusing the dashboard's query layer
- Featured: My Summer as a Platform Engineering Intern at Brain Co.
Undergraduate Researcher at UCLA Programmable Software Systems Lab
- Location: Los Angeles, California, United States
- Period: Feb 2026 - Present
- Link: https://angelhof.github.io/programmable-software-systems-lab.html
- Advised by Prof. Konstantinos Kallas, cross-language profile-guided compiler optimization artifact reuse for serverless runtimes
Software Engineering Intern at Harvey
- Location: San Francisco, California, United States
- Period: Jun 2025 - Sep 2025
- Link: https://harvey.ai/
- Embedded Experience Team, joined as part of the first ~60 engineers, agentic workflows product for Harvey's Microsoft Word add-in
- Built Office workflow automation engine with composable Python blocks, React factory rendering, and Word Office.js integration
- Engineered real-time LLM suggestions system with SSE-based queueing, git-style diffs for Office.js, and granular apply/undo
- Built block-based execution engine with type-safe streaming of heterogeneous outputs (suggestions, text, files), workflow schema validation, streaming abort propagation, and Sentry/Datadog instrumentation
- Designed rubric-based SxS LLM evaluations with Legal Researchers and a Streamlit app to benchmark drafting quality
- Featured: Harvey's improved Word experience understands how you work
Software Engineering Intern at Skylow
- Location: Berkeley, California, United States
- Period: Jul 2024 - Sep 2024
- Link: https://skylow.ai/
- Founding Team, joined as part of the first ~5 engineers, RAG-powered character conversations + social simulation engine
- Built a high-performance RAG pipeline with Pinecone vector search, achieving 90%+ retrieval accuracy under 0.2s latency
- Engineered social simulation system with LLM-driven personalities and vector-based memory for 10,000+ virtual characters
- Reduced NSFW content in the social media feed and user data by 98% via batch-processed AWS Rekognition
- Set up Sentry exception monitoring and PostHog event tracking to analyze user session flows and frontend error patterns
Lead Software Engineer / Founder at ShapeShiftOS
- Period: May 2020 - Apr 2024
- Link: https://shapeshiftos.com/
- Open-source award-winning Android-based mobile operating system with a highly customizable user interface and unique features
- Developed features including custom kernel governor, under-screen fingerprint scanner animations and icons, system theming engine based on color-palettes, Google Lens integration in screenshot tool, dynamic TrueType lockscreen clock, upstream contributions to Google Material Components, etc.
- Grew to 160,000+ downloads across 160+ countries organically with no marketing budget
- Recruited talented engineering team consisting of 25+ Qualcomm device engineers and 3 core developers
- Received SourceForge Open Source Excellence Award (top 0.5% of open-source projects)
Awards & Recognition
Undergraduate
- USACO Gold (2026), USA Computing Olympiad
- Kleiner Perkins Engineering Fellow (2025), Kleiner Perkins (1 of 30) — Out of 3,000+ applicants
- Summer '25
- Selection Committee '26
- Upsilon Pi Epsilon Inductee (2025), UCLA Computer Science
- UCLA Merit Scholarships (2024-2028), UCLA
- Stein Engineering Scholarship (2025), Shirley and Walter Wang Computer Science Scholarship (2024)
- William F. Sharpe Technology Fellow (2025), UCLA Alumni
High School
- Linguistics Olympiads (2023), IIIT Hyderabad (Top 25 in India)
- International Linguistics Olympiad (IOL) Finalists' Camp
- Asia Pacific Linguistics Olympiad (APLO) Indian Delegation
- Regional Asian Math Olympiads (2023-2024) (Multiple Medals)
- Gold Medal in Philippines International Mathematical Olympics (World Rank 3)
- Silver Medal in Fermat Vietnam Mathematical Olympiad (World Rank 7)
- Bronze Medals in Southeast Asian Mathematical Olympiad and Hong Kong International Mathematical Olympiad
- Open Source Excellence and Community Leader (2023), SourceForge File Hosting (160,000+ downloads)
- Recognized for ShapeShiftOS being in the top 0.5% of all open-source projects globally by downloads
- XDA Recognized Developer (2021) (Top 300) — Out of 11 million users
- Recognized as the youngest Recognized Developer on the XDA Forums for Android Development (awarded at age 15) as of 2024
- Board Exam Excellence (2022-2024), Central Board of Secondary Education (99.2% in Grade 10, 97% in Grade 12)
- Sahodaya Utkrisht Puraskar (Education Minister's Award) in Grade 12
- State Rank 1 in Grade 10
Standardized Tests
- SAT Perfect Math Score (2023), The College Board (1580) — 800 Math, 780 Verbal
- TOEFL Excellence (2023), ETS (119/120) — Top 1% globally
Education
University of California, Los Angeles
- Degree: BS in Computer Science
- Location: Los Angeles, California, United States
- Period: 2024 - 2028
- GPA: 3.85/4.00 GPA
- Honors: Dean's Honors List
- Courses: CS 1 Freshman Computer Science Seminar (Fall 2024); MATH 32A Calculus of Several Variables (Fall 2024); CS 31 Introduction to Computer Science I (Fall 2024); LING 1 Introduction to Study of Language (Fall 2024); ENG 1ML Introduction to Engineering Design: Machine Learning (Winter 2025); CS 32 Introduction to Computer Science II (Winter 2025); PHYSICS 1A Physics for Scientists and Engineers: Mechanics (Winter 2025); MATH 61 Introduction to Discrete Structures (Winter 2025); CS 35L Software Construction (Spring 2025); CS 33 Introduction to Computer Organization (Spring 2025); EDUC 10 Introduction to Educational Issues and Scholarship (Spring 2025); ENG 1RK Introduction to Engineering Design: Rockets (Spring 2025); MATH 33A Linear Algebra and Applications (Fall 2025); CS M51A Logic Design of Digital Systems (Fall 2025); MATH 32B Calculus of Several Variables (Fall 2025); PHYSICS 1B Physics for Scientists and Engineers: Oscillations, Waves, Electric and Magnetic Fields (Fall 2025); CS 111 Operating Systems Principles (Winter 2026); CS 174A Introduction to Computer Graphics (Winter 2026); CS 188 Special Courses in Computer Science: Introduction to Robotics (Winter 2026); MATH 170E Introduction to Probability and Statistics 1: Probability (Winter 2026); CS 180 Introduction to Algorithms and Complexity (Spring 2026); CS 188 Special Courses in Computer Science: Artificial Intelligence for Human Interaction (Spring 2026); ECE C247B (Graduate) Neural Networks and Deep Learning II (Spring 2026); PHYSICS 1C Physics for Scientists and Engineers: Electrodynamics, Optics, and Special Relativity (Spring 2026)
Brown University
- Degree: Brown Systems Week
- Location: Providence, Rhode Island, United States
- Period: 2026
- Honors: Cohort of 25 participants; sponsored by NSF, Google, and Amazon
- Topics: eBPF and Kernel Extensions, GPU Performance Modeling, Supply-Chain Security, Formal Reasoning for Systems, Rust Privacy Analysis
Shiv Nadar School
- Location: New Delhi, India
- Period: 2021 - 2024
- GPA: 10.0/10.0 CGPA
- Honors: Valedictorian, AP Scholar with Distinction
- Topics: Physics, Chemistry, Mathematics, Computer Science, English
Projects
ShapeShiftOS
- Tech: Java, C, Go, Smali, Make
- Link: https://github.com/shapeshiftos
- Stats: Downloads: 160,000+, Countries: 160+, Stars: 50+, Open-Source Projects (SourceForge): Top 0.5%
- An open-source Android-based mobile operating system for Qualcomm devices that I founded and led from 2020 to 2024. Shipped custom kernel governor, under-screen fingerprint scanner animations and icons, system theming engine based on color-palettes, Google Lens integration in screenshot tool, dynamic TrueType lockscreen clock, and upstreamed code to Google Material Components.
MiniR (Embedded R-Like Programming Language)
- Tech: C++20
- Link: https://github.com/AshutoshSundresh/MiniR-Holly-Hock2
- A from-scratch R-like interpreter and REPL intended for Casio fx-CP400 family of calculators, including a custom lexer, expression parser, and evaluator that support R-style vectors, matrices, data.frames, subsetting, recycling rules, NA/NaN/Inf semantics, and many core builtins. Designed a compact runtime with attribute-aware values and lightweight containers tuned for constrained embedded environments, plus a simple desktop CLI/REPL interface that can also be embedded as a library.
ResumeIDE (LaTeX Desktop Editor)
- Tech: Rust, Tokio, Tauri, TypeScript
- Link: https://github.com/AshutoshSundresh/latex-resume-editor
- A desktop LaTeX resume editor with live PDF preview using React 19 and Tauri 2. Implemented async Rust backend for real-time LaTeX compilation, Monaco Editor integration with syntax highlighting, system requirement detection with guidance, and cross-platform desktop deployment with native performance.
3D First-Person Pac-Man
- Tech: Three.js, TypeScript, Vite
- Link: https://github.com/AshutoshSundresh/three-js-first-person-pacman
- A 3D first-person Pac-Man game with dual-camera rendering (main view + minimap) in a single render pass. Implemented post-processing pipeline with dynamic bloom effects, procedural environment maps for real-time reflections, spatial audio, and ghost AI with pathfinding across multiple behavioral states.
Bibliotheque (Library Management System)
- Tech: PostgreSQL, Supabase, Next.js, TypeScript
- Link: https://github.com/Ashutosh-s-Test-Dumpster/library-management
- A real-time library management platform achieving <50ms search latency. Implemented secure auth with row-level access control, admin dashboard, and concurrent CRUD operations.
Interests
Fashion
I mix-match a lot of different styles. Currently I own stuff (sweaters, jackets, jeans, etc.) from brands like A-Cold-Wall*, We11done, Diesel, Carhartt WIP, Stussy, Bluorng, Yitai, Almost Gods, Bape, Human Made, Derschutze, Fear of God, Marni, etc.
Games
I’ve worked intermittently on RPG Maker XP to build small Pokémon fangames using Ruby, including design, storyboarding, and a lot of custom scripting required for gameplay logic. In addition to several larger, unfinished Pokémon projects, I’ve spent some time in the Game Boy Advance Pokémon modding scene. Outside of Pokémon, I mostly play classic Need for Speed titles such as Need for Speed II SE, Need for Speed III: Hot Pursuit, and Need for Speed: Hot Pursuit 2, and currently hold 10+ world records on speedrun.com for these Need for Speed games. I also love playing Grand Theft Auto: Chinatown Wars both on mobile and on a DS emulator.
Films
Favourite films: Whiplash (2014), The Pianist (2002), Fight Club (1999), Shutter Island (2010), The Green Mile (1999), The Truman Show (1998), The Shawshank Redemption (1994), Parasite (2019)
The same content is available as plain text at /llms.txt.