Success Stories
The Landmark Achievements Behind Our Engine
Our engine is built upon a foundation of real-world success. Each project below represents a
breakthrough in industrial control and a core capability of our technology,pioneered by our founder, Dr. Amit Gupta, in his previous roles.
Success Story 1 : World's First Closed-Loop Reinforcement Learning Breakthrough: Turning SOx Emissions Compliance into Multi-Million Dollar Profit'
THE CHALLENGE
A facility needed to neutralize SOx emissions but faced constantly
changing, unmeasurable variables in its fuel and neutralizing agent, leading to a recurring
penalty of several million dollars/annum
DR. GUPTA’S GROUND BREAKING APPROACH
In his previous role, Dr. Gupta architected a
world-first closed-loop Reinforcement Learning (RL) system that learned to infer real-time process dynamics from the final emissions sensor alone.
THE PROVEN RESULT
Transformed a recurring compliance cost into a new profit center by
automatically using the absolute minimum resources necessary, 24/7.
THE BRIDGE TO OUR ENGINE
This patented ability to master cause-and-effect with limited
data is a core, proven function of the CygnisAI engine.
Success Story 2: A Predictive Engine for Ultimate Furnace Stability
THE CHALLENGE
Large industrial furnaces at a major facility were chronically plagued by
instability from variable fuel gas quality, wasting millions in fuel and reducing the yield of
high-value products.
DR. GUPTA’S GROUND BREAKING APPROACH
While leading industrial AI R&D, Dr. Gupta
developed a proprietary predictive engine that learns a furnace’s unique physics to anticipate
O2 deviations 10-15 minutes into the future by inferring the fuel's real-time energy content.
THE PROVEN RESULT
A proven model for a step-change in performance: Dramatically
reduced fuel costs, increased high-value product yield, and enhanced operational safety by
holding the furnace in its perfect "Goldilocks Zone."
THE BRIDGE TO OUR ENGINE
This proprietary predictive model is not just an inspiration for
our engine—it is a core, foundational component of the CygnisAI offering, ready for
deployment.
Success Story 3: From Coker Chaos to High-Margin Crude Dominance
THE CHALLENGE
At one of the world's most complex refineries, a multi-year, unexplained
crisis in a critical Delayed Coker Unit was costing millions. This chronic instability and the
severe risk of unpredictable coke carryover made it impossible to process cheaper, high-
margin opportunity crudes like Basra Heavy, leaving tens of millions in potential profit on the
table.
DR. GUPTA’S GROUND BREAKING APPROACH
Leading the initiative for the supermajor, Dr.
Gupta first rejected existing assumptions to diagnose the root-cause physics of the chaos.
Proving the variability was predictable, he then personally developed a predictive model that
cracked the code of coke carryover. This model gave operators the unprecedented ability to
foresee and proactively prevent carryover by mastering the complex interplay between
operating variables.
THE PROVEN RESULT
Tens of millions in new annual profit unlocked. By vanquishing the
chronic coke carryover issues, the refinery could now confidently process high volumes of
Basra Heavy and other opportunity crudes, fundamentally changing the economics of the
facility.
THE BRIDGE TO OUR ENGINE
This project is the direct technical blueprint for how CygnisAI
masters input variability. The core logic Dr. Gupta developed to first decode unexplained
chaos and then unlock a new feedstock is now embedded in our Adaptive Inference Engine.
Testimonial
"We operate one of the most complex refineries on the planet, and Basra Heavy was
our biggest headache. The risk of coke carryover always held us back. In his work here, Dr. Gupta's
model was the breakthrough. He gave our operators the predictive confidence to push the limits
safely. He didn't just optimize a process for us; he unlocked an entire feedstock."
— Refinery General Manager, Global Energy Supermajor
Success Story 4: Mastering Combuston with Real-Time Vision AI
THE CHALLENGE
A plant using petroleum coke for fuel suffered from chronic combustion instability caused by inconsistent petcoke particle size—a variable impossible to measure with traditional sensors
DR. GUPTA’S GROUND BREAKING APPROACH
In a prior role, Dr. Gupta deployed a novel computer vision system on a high-speed conveyor belt. The AI captured particle size
distribution in real-time, feeding the data to a control model that proactively adjusted upstream equipment.
THE PROVEN RESULT
Significant fuel savings and stable operations by and ensuring consistently efficient combustion.
THE BRIDGE TO OUR ENGINE
This success in taming extreme physical variability with highspeed data proves capability
