Electricity has a language problem. A meter can measure, a relay can protect, and a computer can calculate. Getting them to understand one another is a separate occupation. Prasanth Gopalakrishnan has spent much of his working life in that conversational gap. It is an unusual place to build a company: somewhere between machinery that has already earned its keep and software impatient to get started.
His route there ran through electronics and communications at the National Institute of Technology, Calicut, then a master's in electrical engineering at Colorado State University. He worked as a software engineer at Robert Bosch in 1995-96 and as a research assistant at Colorado State in 1997-98. The combination matters to the story. Communications, computing and the physical behavior of electrical systems would become the territory of his business.
Kalkitech dates its beginnings to May 1998. Gopalakrishnan returned to India with an interest in smart grids, the idea of adding communication and computer intelligence to electricity networks. He brought his business proposal to friends who contributed money and expertise. College colleagues Sushil Cherian and Sunil V. Mathew became part of that early company story. The technical ambition arrived before the fashionable vocabulary caught up.
A business with a long attention span
There was an expensive test of that ambition. During an early downturn, the company faced a choice between continuing to develop technology and taking the faster returns available from services. It kept the technology focus. The core team shrank from 45 to 15, and pay was cut. The decision survives in the record because Gopalakrishnan described the sacrifice himself. It was a business decision with consequences for people, rather than a neat slogan for a presentation.
By 2011, Kalkitech had reported revenue above ₹22 crore. A ₹2.5 crore loan in 2006 had supported a Kochi campus; later investment from Kerala Venture Capital Fund helped overseas expansion. These are dated milestones, useful for understanding how the company financed its way forward. They also give the early story some texture: friends, a smaller team, borrowed capital, and products that had to find paying customers.
By 2013, annual turnover was reported at nearly $10 million, with half coming from outside India. Gopalakrishnan was selling a proposition that now sounds familiar: use information to get more useful work from electrical infrastructure. The work included collecting meter readings and identifying patterns in electricity consumption. Long before today's AI infrastructure conversation, he was interested in what better visibility could do for the power system.
The equipment gets to stay
The distinctive feature of Kalkitech's approach is coexistence. Older control systems and newer applications have to operate together. A utility does not acquire an entirely new history when it buys software. Gopalakrishnan's business has made room for the equipment already installed, translating between communications protocols and helping data move between field devices and the systems that use it.
That approach gives an unglamorous word, interoperability, a fairly practical meaning. The purchase of a new application should not automatically turn every existing device into an antique. Kalkitech's work spans protocol implementation, engineering, testing and training. Its tools sit at the points where otherwise useful systems struggle to exchange information. Think of a dinner party whose guests are all competent, but have arrived speaking different languages.
The acquisition of California-based Applied Systems Engineering extended that work. Announced in January 2015, it brought protocol testing equipment and a team with decades of utility communications experience into the business. Gopalakrishnan emphasized the shared commitment to communications products. ASE's repertoire included test sets, translators and alarm annunciators. The attraction was technical continuity: more ways to connect equipment, and more ways to check that the connection actually worked.
A contrarian with a wiring diagram
His own essays offer a more individual view of the founder than a list of product releases. In September 2019, he challenged the assumption that industrial devices at the edge of a network should keep taking on more processing. For many applications, he proposed simpler field devices and more protocol processing in the cloud. He acknowledged that some applications did require substantial local computation.
His concern was the meaning of the data. Reducing industrial information to simple name-and-value pairs could discard context and the model behind it, he argued. Move the raw stream securely, and software elsewhere could interpret it in different ways. This was a specific architectural argument, with assumptions about bandwidth, computing cost and security. It reveals a founder willing to question where the industry puts the work.
Another essay that month asked what artificial intelligence projects were actually supposed to accomplish. Gopalakrishnan criticized initiatives without a clear use case, return on investment, or route from a proof of concept into production. His proposed hybrid approach connected field data, existing enterprise systems and cloud analytics. The impatience was directed at projects that collected technology names faster than they established a practical objective. Even a fashionable acronym, apparently, ought to earn its electricity.
“The business was always there, but people questioned the need for investing in it”Prasanth Gopalakrishnan · 2013
The crowd in Salt Lake City
There is a pleasingly physical moment in his 2019 writing. At Solar Power International in Salt Lake City, Gopalakrishnan noticed the density of the exhibition crowds. Slow down to look at a phone, he observed, and a collision with another attendee became a real possibility. After the cloud architectures and protocol discussions, here was a simpler form of network congestion.
He was struck by the presence of familiar automation businesses, including ABB, Siemens, Eaton and GE, repositioning themselves for solar. He also noted the Chinese exhibitors and the breadth of offerings around solar and storage. His interest went beyond the hardware on the stands. Developers, consultants, software companies and integrators were forming an industry around it. He described the visit as opening his eyes to a different ecosystem.
That reaction makes a useful companion to his longer-range speculation. In another essay, he considered whether solar generation, storage and connectivity might let communities organize with greater energy autonomy. He imagined changes in where people could live and work, and in the services around which communities formed. These were possibilities he was exploring, rather than a prediction with a delivery date. His subject could expand from a protocol to the organization of a town.
More connections, more responsibility
Gopalakrishnan's enthusiasm for distributed energy came with detailed questions about who would be allowed to control it. In an August 2019 essay on cybersecurity, he examined three parts of the system: the energy devices, the networks and cloud services, and utility or aggregator systems. He advocated authentication, encrypted communication and carefully limited permissions. A rooftop installation belonged to a larger conversation once it could exchange commands and data.
His writing on smart inverters made the maintenance problem concrete. Settings would need to change, software would need updates, and utilities would need reliable records of who had done what. Relying entirely on each manufacturer's separate arrangements could restrict choice, he argued. Secure access across equipment from multiple suppliers was both a technical concern and a question of who retained room to act.
This interest in shared rules also appears in his professional commitments. His career includes work with standards organizations and utility industry groups. IEEE working-group meeting records list him with Kalkitech/ASE. At RE+ in September 2023, his presentation concerned telemetry gateways for distributed energy resources and practical experience with Rule 21 integrations. It was a short, specific subject for someone whose business depends on making the connection hold.
- 1998
- Kalkitech begins
- 2019
- Essays on cloud, security and energy autonomy
- 2026
- Data-center power flexibility paper
The same question reaches the data center
The newer setting is the AI data center. A white paper published on Kalkitech's website in January 2026, bearing Gopalakrishnan's name, examines the gap between computing systems and the electrical infrastructure supplying them. He identifies several related problems: delayed grid connections, limited visibility into changing loads, unused computing capacity, and difficulty coordinating IT systems with power equipment.
His proposed response brings utility communications into the management of computing workloads. Some work can wait; other work has immediate service commitments. His paper considers how scheduling, available power and grid signals might be coordinated, while treating latency and completion time as constraints. It is an architectural proposal, rather than a claim that every data center has already achieved the savings it discusses.
The thread through his career is unusually consistent. Earlier, the task was to help utility equipment exchange information. Now he is considering how electrical and computing systems can make decisions with a better view of one another. The setting has changed from meters and substations to AI workloads, but the underlying fascination remains recognizable: capacity can be present and still be difficult to use.
Kalki.io, the company's data platform, gives that interest a commercial form. It collects field information, supports device management and remote access, and connects operational equipment to applications. Available as a hosted service or for deployment on a customer's infrastructure, it reflects the business Gopalakrishnan has kept building: a meeting place for systems that have different origins, different owners and different working lives.
Away from that work, he lists reading and time with family among his interests. The scale comes back down pleasantly after the discussion of grids and data centers. His professional life, meanwhile, keeps returning to the awkward middle of things. The old equipment is still here. The new software has arrived. Someone has to introduce them, and stay long enough to see whether the conversation makes sense.