A warehouse door makes an unusually good place to begin a second act. On one side, the assets can be counted, tagged and located. On the other, they enter the wider world, where the information has a habit of arriving after the goods should have. Sanjay Sharma had helped build a company for the first side of that door. The second side kept him interested.
His earlier venture, KeyTone Technologies, used radio-frequency identification to track assets within an enterprise. After its acquisition, Sharma looked further along the supply chain. Companies could see what they owned inside their premises, yet had much less timely knowledge once it moved between them. The journey was full of handoffs. The information was full of holes. He began asking customers about that gap and found it extended well beyond any one company or country.
Roambee grew out of that question. Today, as co-founder, chairman and CEO of the business renamed Decklar, Sharma is asking a further one. A system can tell you where a shipment has stopped. Someone still has to decide whether to worry, whom to call and what to change. His career has become a succession of attempts to shorten the distance between an observation and a useful response.
The first journey was his own
In 1991, Sharma arrived at South Dakota State University on his first international trip. He had completed his undergraduate studies at the University of Bombay. Two years later, in spring 1993, he graduated from South Dakota State with a master’s degree in electrical engineering. The route from Bombay to Brookings came before the years of Silicon Valley company building.
When he returned to campus in October 2022, he met university and community leaders and discussed connections between the university and industry. “It’s good to be back,” he said. The visit included conversations about partnerships and pathways linking South Dakota State with Silicon Valley. For a founder whose work concerns connections between distant places, returning to the place that began his own international journey was a pleasingly literal exercise.
His engineering career included roles at NASA, EMC and Schlumberger. He also co-founded Plexus Technologies, which became an ICICI Ventures portfolio company, and KeyTone, acquired by Global Asset Tracking Ltd. These were businesses and technical environments with different demands. By the time he was building Roambee, he had experience of both making technology and finding an organization willing to use it.
A sensor needs a business model
Sharma’s account of the early mistakes is refreshingly specific. The business model produced unpredictable sales cycles and deal sizes. Product architecture and the way the company hoped to make money did not fit together. He also came to distinguish the team needed to start a company from the team needed to scale one. A working product, in this telling, was only part of the job.
Building hardware brought its own lessons in expertise, infrastructure and testing. He has pointed to The Hard Thing About Hard Things as an influence, alongside the films Forrest Gump and Jerry Maguire. Persistence and conviction are the qualities he takes from them. His reading and viewing list has room for both venture-capital difficulty and Tom Hanks. Entrepreneurship does occasionally require a broader syllabus than the pitch deck suggests.
“Every data point has an expiration date!”
Sanjay Sharma
The physical proposition became easier to understand through the company’s partnerships. In 2017, T-Systems and Roambee offered a worldwide monitoring service combining portable radio sensors, cloud software and analysis. The devices were called Bees. T-Systems supplied cloud infrastructure; Roambee supplied its application and sensing technology. A small object riding with a shipment depended on a rather large arrangement behind it.
Later that year, Roambee introduced BeeBeacon, using Bluetooth sensors to relay package-level readings through a central Bee. Location alone left unanswered questions about humidity, temperature and handling. The company also offered sensors through a subscription arrangement. Sharma was working on the economics of adoption as well as the engineering: making the unit small enough to follow a package mattered, and so did making the purchase manageable.
In 2019, T-Mobile brought BeeAware to its US narrowband-IoT network. Sharma described four barriers he wanted to remove: cost, complicated setup, ongoing maintenance and data-security complexity. It is a useful list because none of those objections disappears when a demonstration works. Buyers have to live with the device long after the people giving the presentation have left.
Give the person nearest the problem a say
In a 2018 conversation about Indonesian industry, Sharma emphasized training, automation, visibility, analytics and adoption. He wanted analytical insight to reach people throughout the value chain, including those closest to the work. He described automation in terms of transparency, collaboration and quality. Technology adoption in a price-sensitive market had to justify itself through value.
That interest in the people using a system also appears in his career advice. He has encouraged students to combine technical skills with storytelling, design and business understanding, and to work on real problems early. His example is interns progressing from firmware debugging to supply-chain assignments. The underlying advice is practical: become capable of understanding the problem around the code.
For his own organization, he uses the phrase “Responsible Rebels.” It has the advantage of sounding slightly unsuitable for an employee handbook. The argument underneath is orderly enough. People should question a process, understand its purpose and change it when it obstructs the objective. They should also take responsibility for finishing the work.
In his account of the culture, individuals and teams can set goals that align with the organization’s. Managers have individual performance measures alongside team measures, an arrangement intended to encourage mentoring. Asking why is permitted until the reasoning makes sense. Rebellion, here, comes with an obligation to bring colleagues along. A lone objection does not deliver a shipment.
The groceries provide a useful test
A founder can explain a philosophy indefinitely. Tesco offers a more concrete way to examine Sharma’s work. In November 2024, the companies described a deployment covering more than 23,000 container journeys and 6.21 million miles across the retailer’s road and rail network. Solar-powered sensors traveled on its UK containers.
Tesco had previously depended on third-party information to follow those movements. Direct location data helped it confirm deliveries, investigate missing shipments and maintain stock accuracy. Integration with scheduling systems connected the readings to arrivals, departures, dwell times and expected arrival times. Store staff had fewer reasons to ring around departments looking for their goods.
The appeal of this example is its lack of ceremony. Groceries have to reach stores. Containers have to return to use. Missing information generates phone calls and uncertainty. Sharma’s technology becomes meaningful when it changes those routines. The most useful feature may be the conversation a store colleague no longer needs to have.
The new name asks for an answer
On September 10, 2025, Roambee announced its change to Decklar, a name drawn from “Decision Clarity.” Sharma’s customer letter described a broader mission joining visibility, intelligence and real-time decisions. Existing access, contracts and support arrangements would continue. Behind the new name was a familiar commercial task: explain what has changed without making a customer’s working day more complicated.
The ambition extended to replenishment, goods receipt, security and forecasting. A decade of visibility data would support recommendations about the next action. That shifts the promise of the product. A location record can be accurate and still leave the operator with a queue of unresolved choices. The company now wanted to help with those choices too.
In May 2026, Sharma set out his objection to applying general-purpose AI agents to logistics without rebuilding the supporting systems. Physical operations impose deadlines. Information arrives through conflicting feeds, with different identifiers and timestamps. The system must recognize which shipment, contract and customer an event concerns before choosing a response. His argument puts much of the difficult work beneath the conversational interface.
He also called for decisions that can be explained afterward, and for connections into the operational software where actions actually take effect. A fluent recommendation is a limited achievement if nobody can execute it. For an engineer who began with the whereabouts of physical things, the emphasis on a verifiable chain of events is consistent.
There is still a team to coach
By July 2026, Sharma’s update described RADAR executing workflows across dispatch, transit and delivery. Decklar reported a 60–70% reduction in alert response time across 11 Global 2000 deployments. These are company-reported results. His stated next goal was to extend autonomous resolution from individual workflows to entire networks.
The same update offered a smaller experiment: engineering staff used the company’s DINA voice AI to collect progress reports and summarize blockers. Other functions adopted it. The company was applying its own tools to the ordinary business of getting colleagues updated. Even a supply-chain AI company must contend with the meeting calendar.
Outside work, Sharma coaches cricket and mentors young professionals. Coaching is an apt companion to the management philosophy he describes: people need the reasoning, the skills and enough room to act. His return to South Dakota State belongs to that same part of the story, where experience becomes something to pass along.
The warehouse door remains a useful image for his career. Each solution has exposed a question just beyond it. After locating the asset came understanding its journey. After understanding came deciding what to do. Sharma is still following the goods, and increasingly the decisions that determine whether they arrive as intended.