​A technical evening for engineers working at the intersection of AI, optimization, and regulated domains.

​A 40-minute technical talk by Maprang Suwanbubpa, founder of Checkmate AI, on the engineering problems behind GREYZONE™ — a decision-support system for race teams making regulatory exploit allocation decisions across a season.

Check out MVP Demo for GREYZONE™ Formula 1

​
The talk goes deep on three things:

​Why F1 and SUPER GT are different engineering problems, not different deployments of the same product.

​F1's constraint surface is dominated by aero development economics — cost cap, wind tunnel hours, homologation cycles.

SUPER GT's is dominated by season-handicap dynamics where every championship point adds 2 kg of ballast to the next round, with regime changes at 50, 67, and 84 kg that introduce fundamentally different penalty mechanisms.

​Why finding regulatory loopholes is commoditized, and the constraint layer downstream is the actual work.

​A recent validation finding with a former SARD engineer reshaped how Checkmate AI thinks about the company. 15 candidate ambiguities. 12 already known to teams. The reasons they're not pursued aren't about discovery — they're about sponsor optics, driver workload, championship position, and authority discretion. The talk walks through what this means for AI tools in expert-heavy domains generally.

​Three specific open problems

​Checkmate AI is about to start working on, presented honestly as "this is the work ahead, here's why it's hard, here's the kind of engineer who'd think well about it":

• ​Multi-agent path-dependent optimization with non-stationary penalty functions

• ​Bilingual regulatory parsing where translation cannot be a preprocessing step

• ​Constraint elicitation interfaces for tacit team knowledge under privacy constraints

​The talk is followed by extended Q&A and networking. For engineers who hear one of the three open problems and think "I know how to start this" — there's a sign-up sheet at the back, and Maprang will personally follow up with the sharpest questions.

​Who this is for

• ​Principal engineers and software architects working on hard applied-AI problems

• ​People with multi-objective optimization, applied NLP, or domain-expert tooling backgrounds

• ​Engineers from finance, sports analytics, healthcare AI, legal tech, compliance, or any field where AI meets tacit expert knowledge

• ​Former or current F1 or SUPER GT motorsport technical staff

• ​AI/ML engineers curious about constraint modeling in regulated domains

• ​Founders working at the intersection of vertical AI and domain expertise

​Honest filter: this is a small-company technical recruiting event. If you're early-career or looking for AI-feature-shipping work, this isn't the right event.

​If you want to hear about a meaty multi-year engineering problem in a domain that doesn't have ten startups chasing it, register.

​Format

• ​6:00 PM – 6:30 PM — Doors, drinks, networking

• ​6:30 PM – 7:10 PM — Technical talk (~40 min)

• ​7:10 PM – 7:45 PM — Open Q&A and discussion

• ​7:45 PM – 8:00 PM — Sign-up sheet, structured networking, close

​About the speaker

​Maprang Suwanbubpa is the founder and CEO of Checkmate AI, a Tokyo-based startup building decision-support systems for race teams.

​Previously:
• EMEA Solutions Engineer at Immutable,
• Consultant Engineer at Capgemini Engineering working on predictive maintenance algorithms for Airbus,
• Digital Business Integration Analyst at Accenture,
• Software Engineer at Thomson Reuters.
• B.A. German & Computer Science, University of British Columbia.

​Checkmate AI's first event in this series, hosted at Google for Startups Shibuya on May 19, 2026, gathered close to 200 attendees and panelists from McLaren Japan, Red Bull, Netflix Drive to Survive, JAF, TOM'S Racing, and Track Titan.
See photos on LinkedIn here >> https://www.linkedin.com/posts/cm8ai_futureofspeed-checkmateai-greyzone-activity-7460993145426677762-v_aN?utm_source=share&utm_medium=member_desktop&rcm=ACoAAEJT-RcBofA8TF32mtCn_AF6mFSQqgIK96w

​About Checkmate AI

​Building GREYZONE™ — decision-support for regulatory exploit allocation in motorsport. F1 MVP shipped.

​SUPER GT in active definition with former GT500 engineering validation.

​Multi-year roadmap extending to WEC, IndyCar, MotoGP, and eventually aerospace certification environments.

​Pre-seed round in progress. check-mate.ai

​Thanks to Tech Hub Yokohama and Mitsubishi Estate for the venue partnership.

​