About
Hi — I'm Arshil, a software engineer working across game development, machine learning, full-stack, and robotics. I tend to start by diving straight at a problem with whatever skills I already have, then level up on whatever the problem turns out to need. That habit is why the work below spans four disciplines instead of one.
Right now I'm a research student at Hiroshima University, studying memory optimisation techniques for 3D Gaussian Splatting (3DGS). Alongside that I run MakeSense solo — an AI learning tool with 50+ users and 100K+ organic content views, built and shipped in two weeks.
Before that: an internal AI “Marketing OS” at Envesty, still in use after the internship ended, and two game-programming internships spent on object pooling, physics systems, and runtime performance. Four national hackathon placements along the way, each one built inside 48 hours.
Experience
- 2026 — Present
Founder · MakeSense
Validated the concept before writing product code: ran 10 ad variants on a ₹1,000 total budget, and the eventual MakeSense concept generated 4x more leads than the average variant. Built and shipped the first version in two weeks, and now run product, content, and distribution solo. Grew to 50+ users and 200+ weekly site visits, driving 100K+ organic content views without paid spend.
- Next.js
- React
- Product
- Growth
- 2026
Software Engineer · Envesty
Built an internal AI-driven “Marketing OS” — a Claude-based system modeled on MakeSense's own — that tracks the founder's past posts and current trends, then generates ready-to-use content cycles and scripts on request. Added a Reddit trend-pull so it could surface real posts as inspiration, streamlined coordination with a Notion + Slack integration, and shipped editable design templates the founder could extend himself. Still in active use at the company post-internship.
- Claude API
- Notion API
- Slack API
- Python
- 2024
Game Programmer Intern · Singular Scheme
Pioneered high-impact performance optimization systems, including an object pooling architecture that cut runtime overhead. Conducted system analysis and optimization work that directly improved game stability and responsiveness.
- Unity
- C#
- Performance
- 2023
Game Programmer Intern · IDZ Digital Limited
Led development of 10+ technical prototypes in Unity and C#, rapidly validating novel game mechanics and accelerating R&D cycles. Engineered physics simulation systems and optimized algorithms for runtime efficiency in complex game environments.
- Unity
- C#
- Physics
Awards
- 2026
National 1st, Federated Healthcare Hackathon
IIT Kanpur, CDIS, NHA, ICMR, DHR
- 2025
Global Community Favorite
Google Maps Platform
- 2025
Semi-Finalist, Handloom Hackathon
IIT Delhi
- 2025
Runner-up, CyberCup 5.0
Amity University Noida
- 2024
Runner-up, Game Jam
BITS Pilani
- 2025
Runner-up, Hackemon
SheBuilds, GeeksRoom
Projects
MakeSense
50+ users, 200+ weekly site visits, 100K+ organic content views, zero paid spend
An AI learning tool that maps hard concepts onto fandoms you already know. Validated the idea before writing a line of code: ran 10 ad variants on a ₹1,000 budget, and the winning one pulled 4x more leads than the rest.
- Next.js
- React
ZenDrive
Community Favorite at the Google Maps Platform Awards
A multi-model system preventing drowsy-driver deaths, with an EMG sensor at its centre.
- EMG Sensor
- Arduino
- React Native
- Expo
- Express
- Node.js
- MongoDB
- Python
- scikit-learn
- NumPy
- Pandas
- Matplotlib
- YOLO
- Google Maps
Inwards
Solo entry for Epic MegaJam 2025
Emotion-driven mechanics — Frustration and Regret — built solo in Unreal Engine 5, exploring resilience after repeated failure.
- Unreal Engine 5
Flood Simulation
Placeholder — details to come
Awaiting details. This card is wired and styled; only the copy, link, and tech list are missing.