fig. 1 — me, from one photo · 42,000 pts

Looking for Summer 2027 internshipsWest Lafayette, IN ·


I build software that learns, sees and ships — from evolved race-car drivers to underwater 3D vision.

Initialize network↓
01 — Architecture

argmax(Tanay)

Hover a node to trace its path. Click to open it.
19 nodes · 40 edgesepoch 000 · loss 0.900
InputHidden · ExperienceHidden · ProjectsFeaturesOutputPurdue AI + CSIndustryBias / GritApplePantheonIEEE ROVR&D Venture StudioBoundary RSSRacing GenerationsTanayOSROV VisionPaper TraderAutonomous MotorsportsEngine Data GANSolar ScoreVision & MLSystems & BackendSimulation & 3Dargmax(Tanay)
Neural architecture v2.0 // 5-layer topology← drag · tap a node
02 — Selected work

Things I've shipped

Racing Generations screenshot

01 / 04 — Personal

Racing Generations

Live demo

Neural-network drivers that teach themselves to race — from a 2D sim to full 3D TORCS.

90%
Held-out tracks finished
46.9s
TORCS lap, gen 157
100
Cars per generation
How it works→
TanayOS screenshot

02 / 04 — Personal

TanayOS

Live demo

My personal AI command center — one chat box that routes to coding agents, local models and my calendar.

373
Tests
9
Integrations
53k
Lines of code
How it works→
ROV Vision screenshot

03 / 04 — Club · Purdue IEEE ROV

ROV Vision

Live demo

Turning underwater photos into 3D models you can measure — for Purdue's competition robot.

92.6%
Detection accuracy
−18%
Depth error
How it works→
Pantheon screenshot

04 / 04 — Company

Pantheon

Site tour

Competitive extracurricular programs for K-12 schools — debate, DECA, Model UN and more. I'm the CTO.

How it works→
03 — About
Portrait of Tanay Jain
Studying
B.S. Artificial Intelligence & Computer Science
Minor
Business Economics
Graduating
May 2028
GPA
3.73

Hi, I'm Tanay.

I'm a junior at Purdue studying Artificial Intelligence and Computer Science, with a minor in Business Economics. This summer I was a software engineering intern at Apple, working on AI video generation for an internal platform with 200K+ users.

Outside class I'm the Vice President of Purdue IEEE ROV, where I work on the vision stack for our underwater robot, and the CTO of Pantheon, which runs competitive extracurricular programs for K-12 schools.

I'm from Fremont, CA. I'm happiest when a project has a number I can push — accuracy, lap time, latency — and a demo I can show someone.

Experience

  1. Summer 2026

    Apple — Software Engineer Intern

    Sunnyvale, CA

    • Built and productionized scalable AI video generation, making end-to-end workflows 4× faster on an internal platform serving 200K+ users.
    • Integrated 15+ HeyGen APIs and 20 generation controls across avatars, voices, rendering and video quality.
    • Worked with HeyGen to define Apple's API requirements — evaluated 12 capabilities and helped shape their roadmap.
  2. 2024 — Now

    Pantheon — Chief Technology Officer

    pantheonprep.com

    • Lead the technical side of a company that runs debate, DECA, FBLA, Model UN, Science Olympiad and chess programs for K-12 schools.
    • Own the web platform end to end and the AI-powered learning tools built on it.
  3. 2024 — Now

    Purdue IEEE ROV — Vice President + Software Engineer

    West Lafayette, IN

    • Built a real-time underwater object detection pipeline reaching 92.6% accuracy with OpenCV and TensorFlow.
    • Engineered a stereo vision algorithm that cut depth-estimation error by 18%, turning depth maps into 3D point clouds with Open3D.
    • Created a 360° photosphere stitching method for immersive underwater visualizations.
  4. 2025 — 2026

    R&D Venture Studio — Software Engineer

    Remote

    • Built an end-to-end e-commerce automation platform on Supabase, containerized services and third-party APIs.
    • Implemented order orchestration: inventory sync, supplier selection and delivery tracking, with analytics on every stage.
  5. Summer 2025

    Boundary RSS — Software Engineer Intern

    Remote

    • Trained CNNs on 5TB+ of seismic data to detect hydrocarbon boundaries, raising detection accuracy by 23%.
    • Applied 3D viscoelastic modeling and Kirchhoff migration, speeding up depth conversion by 38.8%.

Toolbox

Languages
Python · TypeScript · Java · C / C++ · C# · SQL · R · MATLAB
AI & Data
PyTorch · TensorFlow · OpenCV · scikit-learn · NumPy · pandas · RAG · Multi-agent systems
Systems
React · Next.js · Node · FastAPI · Spring Boot · AWS · Supabase · Docker · Unity
04 — Contact

Let's build
something.

Looking for Summer 2027 internships. Email is the fastest way to reach me.