# Imubit

> Houston-based Imubit sells industrial AI software that models complex process plants, tests operating strategies in simulation, and can write optimized setpoints into existing control systems. Its pitch is unusually concrete for enterprise AI: use years of plant data to improve yield, energy use, and margins without building new physical capacity. The hard part is not the neural network; it is earning permission from operators to close the loop.

- **Founded:** 2016
- **Headquarters:** Houston, United States
- **Founders:** Gilad (Gil) Cohen (CEO & Co-founder), Nadav Cohen (President, CTO & Co-founder), Meir Tseitlin (Co-founder, Experimental Technologies)
- **Team size:** Approximately 200 employees in the supplied dataset; third-party public profiles place the company in the 101-250 range.
- **Products:** Deep Learning Process Control (DLPC), Industrial AI Platform, Optimizing Brain, Foundation Process Model, Industrial Solution Applications
- **Notable:** Raised $50 million in disclosed funding by August 2021., Reported adoption by 7 of the top 10 U.S. refiners and more than 90 applications by early 2025., Reported more than 100 closed-loop applications and over ten patents by 2026.

## Products & services

- **Deep Learning Process Control (DLPC)** — An on-premises control application that uses neural models and reinforcement learning to optimize process setpoints in open- or closed-loop modes through existing plant systems.
- **Industrial AI Platform** — A cloud environment where plant teams build, evaluate, simulate and monitor process models and their performance.
- **Optimizing Brain** — A combined solution built around a plant-specific Foundation Process Model and reinforcement-learning controller for coordinated optimization.
- **Foundation Process Model** — A reusable, continuously refreshed model trained on a plant’s process history and engineering knowledge for control, simulation, training and planning use cases.
- **Industrial Solution Applications** — Applications including operator optimization training, planning-tool augmentation, process degradation insight, dynamic simulation and performance dashboards.

## Achievements

- Raised $50 million in disclosed funding by August 2021.
- Reported adoption by 7 of the top 10 U.S. refiners and more than 90 applications by early 2025.
- Reported more than 100 closed-loop applications and over ten patents by 2026.
- ARC Advisory Group identified Imubit as a leader in the emerging Closed Loop AI Optimization category in 2024.
- A hydrocracker case study reported a $20,000-per-day increase in economic objective value and payback inside six months.
- An Oxbow application reported 1-3% higher yield and 15-30% lower natural-gas consumption, with closed-loop implementation in six months.
- A Marathon Petroleum coker application reduced suboptimal cycles and associated giveaway by 25%.
- Opened an R&D center in Cluj, Romania, in January 2025.

## Latest updates

- **2024-09** — Launched the reinforcement-learning-powered Optimizing Brain and Industrial AI Platform around the Foundation Process Model concept.
- **2024-10** — ARC Advisory Group placed Imubit as a leader in a new Closed Loop AI Optimization market category.
- **2025-01** — Opened an R&D center in Cluj, Romania, focused on customer-built and customer-managed industrial AI models.
- **2025-02** — Announced a partnership with Preem for FCC optimization at the Lysekil refinery in Sweden.
- **2025-10** — Published an energy-first operating thesis that frames emissions reduction as a cost and competitiveness lever.
- **2026-04** — Began an open-loop AI pilot with Evonik at its Singapore facility.
- **2026-05** — Partnered with Ketjen on real-time catalyst intelligence, with pilot deployments underway in North America and Europe.

## Links

- Website: https://imubit.com
- LinkedIn: https://www.linkedin.com/company/imubit/
- Twitter/X: https://twitter.com/imubit1?lang=en
- YouTube: https://www.youtube.com/@Imubit/videos

---

Profile page: https://yespress.io/imubit
Published by YesPress — https://yespress.io
Last updated: 2026-08-21