About us

About AdeptDC

Software to Align Cooling Supply with Demand

Every data center is overcooled. Operators blast cold air to get rid of insidious hot spots, which slow down IT operations and increase energy bills. The irony is that even with overcooling data centers are still plagued by hot spots.

AdeptDC’s award-winning off-the-shelf software app, Valency, cleanses hot spots from data centers by connecting IT operations to cooling controls. It works with any type of IT/cooling hardware, floor layout, and data center configuration.

Our Mission

We are living in a digital world, run by data centers. The sustainability of this digital world is being threatened by the ever-increasing demand for operational resources—such as electricity, water—in data centers. The obstacle to solving this data center sustainability challenge is posed by the discordant interactions of the different constituent subsystems such as software stacks, IT hardware, precision cooling systems, and power delivery infrastructure. In that context, the misalignment between cooling demand, posed by IT operations, and cooling supply causes undesirable hot spots and eventually exacerbates into unplanned service downtime, inefficient cooling operations, high maintenance costs, high equipment failure rate, and low staff productivity. This problem of IT/cooling misalignment has caused $30B waste in 2016. Our mission is to solve this problem of IT/cooling misalignment.

Team

Rajat Ghosh, PhD

Founder/ CEO

Prem Talreja

Chief Revenue Officer

Sachin Naik

Senior Software Engineer

Vishal Kumar

Frontend Developer

Raunaq Gupta

Designer

Investors

Ankur Jain

Managing Partner

Alok Nandan

Partner

Advisors

Digvijay Singh, PhD

Chris Johnston

Our Values

As an organization, we value meritocracy with following guiding principles

Self-expression

Be yourself. Everybody else is already taken.

Scientific Thinking

We emphasize scientific thinking over analogy-based thinking.

Accountability

We match authority and responsibility.

Process-orientation

We aim to install a robust system that wins consistently.

Empathy

We work to make changes that we like to see in the world.

Transparency

We aspire to be in harmony with what we think, what we say, and what we do.

Sustainability

Sustainability means striking the rightbalance between today’s urgency and tomorrow’s need.

Get in Touch

I agree to receive information on AdeptDC products and technology

In the absence of an appropriate pattern recognition algorithm, granular and real-time data is just computing burden. To fully utilize it, it is important to be supported by an efficient pattern-recognition algorithm.
Enterprise-grade servers supporting AI workloads are at least few orders of magnitude more expensive than the traditional volume servers. The equipment failure cost is too high to be ignored.
Typical lifetime of an IT equipment is 6-18 months, while that for a facilities equipment such as a Liebert cooling systems 15-20 years. The deployment times are also very different. Furthermore, the typical time-scale for data center facilities operations is in order of few minutes, while that for IT systems is few minutes. Therefore, matching data center facilities to its IT demand is very difficult.
Following Moore law’s, the number of transistors in electronic circuit is rapidly increasing at 1.3x annual rate. Moreover, virtualization of computing resources made data center workload rapidly non-uniform. This two factors lead to a complex data center operation.
Data centers are increasingly supporting application ecosystem such as video streaming, high-frequency merchandise, and real-time image recognition. This new generation applications require fast and high-throughput computation.
A typical data center hosts a multitude of IT, power, and cooling systems with varied response characteristics. This makes data center thermal profile non-uniform and complex. In addition to that, turbulent air cooling adds to the data center thermal profile to an extent that it become non-linear. Therefore, it is literally impossible to build a traditional formula-based mathematical modeling for data center thermal evolution. It necessitates a data-driven real-time framework for data center thermal modeling and prediction.
As much as 50% of data center downtimes happen due to thermal issues. AdeptDC’s data-informed predictive maintenance and automated cooling optimization reduces the risks of these downtimes.
Most of the hardware failures occur due to temperature, voltage, power, or humidity related issues. AdetpDC’s solution reduces the hardware mortality rates with its granular and real-time monitoring and end-to-end feedback control solution.
With data center cooling demand being difficult to predict, data center managers are forced to over-compensate their cooling capacity. AdeptDC’s end-to-end automated cooling optimization removes guessworks from the cooling controls, making it more efficient.
Although most of the data center managers are trained IT professionals, a significant part of their work hours are spent on calibrating cooling system and fixing thermal issues. AdeptDC’s predictive maintenance application intends to save time for IT professionals so that they can focus more on their core jobs.
Data center maintenance could be an intractable problem with unpredictable nature of data center operations and complex thermal response patterns. AdeptDC aims to removes this uncertainty and provides data center managers with much needed peace of mind.