Our story

Built in Pune.
For the strip.

MotoQuant started as a single question at a drag strip outside Pune: “How do I know which part will actually move my ET?” Every tuner, racer, and weekend warrior we talked to had the same problem. The answers were guesswork, YouTube comments, or ₹50,000 test runs. There had to be a better way.

The Problem We Solve

Indian drag racing has exploded over the last decade. Aamby Valley, MMRT, BIC, Kari, Hyderabad — the strips are full. But the tooling hasn't kept up. Racers still tune by feel and magazine estimates. Tuning shops quote improvements from memory. Parts distributors can't tell you how much faster that exhaust will actually make you.

Meanwhile, the global sim tools that do exist are built for international bikes and international conditions. A tool calibrated for Santa Pod at 5°C doesn't tell you anything useful about Aamby Valley at 35°C and 1100 m density altitude in November.

MotoQuant is built specifically for the Indian drag scene — with Indian bikes (all 57 of the most common Indian-market machines), Indian parts prices in INR, Indian venues with real seasonal weather data, and Indian community benchmarks used for validation.

Why First Principles?

We could have built a regression. Train on a database of published ETs, predict new ones by interpolation. Fast to build, reasonable accuracy for common bikes.

But regressions break silently. They can't tell you why your R15 V3 is a second slower than predicted. They can't model the interaction between a new clutch kit and a heavier flywheel. They can't simulate what happens when you run the same bike at 5°C versus 38°C.

First-principles physics can. Every sub-model — Pacejka tire, Willans friction engine, RK4 integrator — produces physically interpretable outputs. When the model is wrong, we can find and fix the cause. When you add a part, the physics of that part propagates through the whole system.

Top-10 mean error after our April 2026 calibration pass: ±0.032s. Reference baselines within 0.001s of targets. That's what first principles buys you.

Founders

Engineers first. Racers second. Shipping a tool we'd use ourselves.

A
Arya Desai
Co-founder · Pune, India

Automobile engineer — M.Tech at NIT Warangal, B.Tech from Manipal Institute of Technology (Automotive System Design). The physics-and-shop-floor side of MotoQuant.

At NIT-W his research runs close to MotoQuant's core stack: a CFD validation study of split-injection strategy in a single-cylinder diesel engine (ANSYS Forte, 28-species n-heptane kinetics, 60° sector mesh, peak-pressure match within 1.6% on the optimal case); a semi-active MR suspension paper that uses six-seed Monte Carlo on ISO 8608 roads to show comfort-optimised skyhook controllers actually lose 18–23% in ride comfort once realistic bump-stops engage — with a full reproducibility artefact; and a design patent filed with NIT-W's IP Cell.

Hands-on side: diagnosed and performance-tuned 25+ luxury and performance vehicles at Power Solutions Pune — Mercedes AMG, BMW, Audi, Jaguar Land Rover, Lincoln — including a flagship Mercedes AMG SLS valvetronic exhaust install. ECU-tuning certified (Techmaghi / ReynLab). Undergrad research at MIT Manipal on PU-based magnetorheological elastomers delivered a 22% improvement in vibration damping.

motoquant@gmail.com
H
Hritwik Mudhai
Co-founder · Manipal, India

The physics + software side. Built the 15-sub-model drag simulation engine, the 287-bike catalog, the ROI knapsack, and the calibration pipeline that fits sim output to real Dragy traces within ±0.05 s.

AI engineer by trade — currently building a separate AI startup in parallel. That expertise is baked into MotoQuant's machine-learning layer: the DNN surrogate (1 ms per ET prediction), the XGBoost-backed ROI ranker, the Bayesian build optimiser, and the natural-language tuning interface.

Wrote the RK4 integrator, the Pacejka tire model, the expansion-chamber 2-stroke curve, and the clutch-slip thermal model from the ground up. Cares more about which terms actually survive the force balance than about chasing a good-looking regression. Wants MotoQuant to be the tool every serious tuner already has open on a second monitor.

dhisetu@gmail.com

Where We're Going

Roadmap — no specific dates, shipping when it's right

Phase 1–2Done

Physics engine + Database

15 sub-models, 287 bikes, 424 parts, 20 venues, 838 tests. Full validation suite.

Phase 3Done

Web launch + Calibration pipeline

FastAPI backend, Next.js frontend live at motoquant.in. Dragy/dyno/ECU import + auto-calibrator.

Phase 4In progress

AI & Optimisation

Parts ROI knapsack, Bayesian build optimiser, XGBoost & DNN surrogates, Smart Recommendations.

Phase 5In progress

Native Desktop App

Tauri 2 + React 19 shell. Offline physics sidecar. macOS, Windows, Linux. Closed beta.

Phase 6Planned

Commercial launch + AI layer

Tiered pricing. B2B Pro Studio for tuning shops. Conversational AI tuning (Mistral 7B). PPO launch control.

Get in Touch

Whether you're a tuning shop interested in the Pro plan, a racer who found a bug, or just want to talk drag racing — reach out.

Location
Pune, Maharashtra, India
Pro licensing
Request a quote →