Inviscid AI

Inviscid AI

Real-time Building Simulations to Optimize Energy and Operational…

Winter 2026ActiveIndustrialsEnergyWarehouse Management TechIoTSustainabilityProptechSan Francisco, CA, USA; Remote
Inviscid AI builds physics-informed AI solutions that transform how buildings and data centers operate. By combining real-time IoT sensor data with computational fluid dynamics (CFD) modeling, we create digital twins that simulate building performance in real time and autonomously optimize operations. Our platform optimizes airflow patterns and ventilation strategies to eliminate dead zones, improve air distribution, and reduce the load on mechanical systems. On the energy side, we minimize HVAC power consumption, reduce cooling costs, and lower overall operational expenses while maintaining optimal thermal comfort and indoor air quality. Beyond immediate operational efficiency, we optimize equipment scheduling and maintenance cycles by predicting system behavior under different conditions, allowing facilities managers to proactively address issues before they become problems. Our physics first approach ensures that we're not just optimizing against historical patterns, but optimizing based on a deep understanding of how air, heat, and energy actually move through your building, enabling us to find solutions that traditional rule-based or purely data-driven systems would miss.

Verdict

High Signal
Market Opportunity
Building energy optimization and HVAC efficiency is a massive market — commercial buildings account for ~40% of US energy consumption and the global building automation market is $100B+. ICP is reasonably clear: facilities managers at commercial buildings and data centers. Monetization path via SaaS + integration services is plausible, and data center cooling optimization is an especially hot subsegment given AI infrastructure buildout.
High Signal
Founder Signal
Kabir Jain: Oxford CS, NeurIPS 2024 paper and ICLR 2025 spotlight paper (top-tier ML publications at undergrad level), Gold medal at International Olympiad in AI (1st internationally in take-home round representing Singapore), Lee Kuan Yew Math and Science Award (top 12 pre-university students in Singapore), MIT JuliaLabs (physics-informed ML for PDE solving), and PUB Singapore (physics-informed storm surge forecasting — literally the same domain as Inviscid's core tech). Qiu Ziming: NUS CS + Mathematics double honors, 5.0/5.0 GPA, Dean's List multiple times, Special Programme in Mathematics. Two exceptional academic founders with directly relevant physics-informed ML expertise built in prior roles.
Medium Signal
Competition
Real incumbents exist: Siemens Enlighted, Johnson Controls OpenBlue, Schneider Electric EcoStruxure, and startups like 75F and Willow. Physics-informed neural networks for CFD acceleration is a genuine technical differentiator vs. pure ML or rule-based BMS systems, but NVIDIA Modulus and Ansys are pursuing the same approach at the simulation layer. The real-time digital twin wedge (milliseconds vs. hours for CFD) is a meaningful moat if it holds at scale, but incumbents have distribution advantages.
High Signal
Product
Named pilot with Kajima (The GEAR by Kajima, Singapore — a real commercial building used as an innovation testbed). Signed government MOU with MPSEDC for a live digital twin deployment at the Madhya Pradesh State Data Center, India, in the presence of the Chief Minister. Customer message (shown publicly) reports 2x cooling efficiency gains. Attending SEMI-THERM 2026 as an exhibitor. These are real named deployments, not internal demos — website metrics (240x faster CFD, 600x faster storm surge) are consistent with pilot results.
OverallA Tier

Inviscid AI is one of the more underrated companies in this batch. Kabir Jain's credentials are exceptional for an undergrad: Gold IOAI, NeurIPS + ICLR spotlight papers, MIT JuliaLabs physics-informed ML, and PUB Singapore storm surge forecasting — the exact problem domain he is now commercializing. Qiu Ziming brings rigorous math/CS depth (NUS 5.0 GPA, double honors). The product has real named customers: a Kajima pilot, a signed government MOU at a state data center in India, and verifiable customer success results. The market (data center cooling, building energy) is large and urgent given AI infrastructure buildout. Competition from incumbents is real but the physics-first approach is a genuine technical edge that rules-based BMS systems cannot replicate. Biggest risks: both founders are still undergrads, go-to-market in enterprise facilities management is slow, and pilot conversions to paid contracts are unconfirmed.

Active Founders

Kabir Jain
Kabir Jain
Founder

I like physics and AI

Ziming Qiu
Ziming Qiu
Founder

I make simulations :).

Inviscid AI
Inviscid AI
TierA Tier
BatchWinter 2026
Team Size2
StatusActive
LocationSan Francisco, CA, USA; Remote
Last Updated3 days ago