A security camera has an excellent memory and terrible manners. It records the lunch rush, the slow queue, the unopened carton, the handoff that took four minutes too long. Then it says nothing. The footage retreats into a hard drive, awaiting an incident dramatic enough to justify rewinding Tuesday. Adit Chhabra looked at that mute archive and saw an unfinished sentence.
His company, Wobot AI, gives the sentence a verb. It connects computer vision to cameras that restaurants, retailers, hotels, and factories already own. A business chooses a task worth noticing. The software watches for the event, tags the relevant moment, and sends an alert or folds it into a dashboard. The proposition is deliberately unglamorous: do not replace the ceiling full of hardware. Make it useful.
That practicality is the through-line in Chhabra’s career. Before he was selling video intelligence, he was studying the anatomy of repeatable work - in finance, acquisitions, corporate integration, hotels, and an earlier software venture built around audits. His subject was never really the camera. It was the troublesome distance between a process written down and a process carried out.
“Be in love with your problem and that is how you will find the right solution.”Adit Chhabra, speaking to Purdue students in 2019
I. The operating educationA degree in how things actually get done
Chhabra graduated from Purdue University’s Krannert School of Management in 2010. His studies covered business, and he also completed Purdue’s entrepreneurship and innovation certificate. The combination offered two lenses he would keep using: how an organization allocates resources, and how a new product earns permission to exist.
The early résumé is a tour through consequences. He worked in capital markets at J.P. Morgan, then joined Amtek Global Technologies, moving through analyst, manager, and director roles in mergers, acquisitions, and integration. He also advised Mapple Hotels Palaces Resorts on business development. Deals promise transformation in a spreadsheet; integration determines whether Monday morning feels any different. Hospitality is even less forgiving. A guest experiences the process, not the policy binder.
An MBA at IE Business School in Madrid followed in 2014 and 2015. Then came Digi Pro Labs, where Chhabra led an AI-enabled platform for monitoring standard operating procedures. Its workflow asked people to document completed tasks, including with photographs. The idea was sensible, but the method carried a small tax: proof depended on the person doing the work remembering to create proof.
The camera nearby had already captured the sequence. It did not need a reminder, a form, or a cheerful notification. It needed interpretation.
II. The pivotKeep the problem. Replace the proof.
In December 2017, Chhabra co-founded Wobot Intelligence with Tapan Dixit and Tanay Dixit. Their central observation was enormous in scale and ordinary in appearance: people across physical industries perform billions of manual tasks, while useful operational data sits behind camera lenses. A camera catches the full motion of work where a checklist captures a claim and a photograph captures a moment.
Wobot began building templates for food service and production, then added retail, hospitality, and manufacturing. The software could look for a defined deviation or pattern and bring it to a manager’s attention. It turned video from a security artifact into a stream of operational evidence.
How a silent feed becomes a decision
- Use an existing camera
- Choose a task to detect
- Tag the relevant event
- Alert a person who can act
There is a product lesson tucked inside the architecture. Enterprise software often arrives with a renovation project attached. Wobot made the installed base its advantage. Existing IP cameras and recorders could stay in place, limiting new equipment and shortening the path to a trial. The visible object remained the same. Its job description changed.
Customers sharpened that job description. Chhabra has said Wobot was built by its customers and continues to be shaped by them. He has also described a company willing to disrupt itself before somebody else does. Those two ideas met in 2020, when the industries Wobot served were forced to rework operations quickly. The team returned to the drawing board.
III. The rebuildA product that had to explain itself
The result, launched in 2021, was a simplified, do-it-yourself SaaS product. Businesses could connect cameras, choose among more than 100 tasks, control schedules and billing, and view feeds across teams. A free tier covered as many as ten cameras. Chhabra called the launch a step toward democratizing the offering.
“Democratizing” is a large word for an onboarding flow, but the mechanics matter. A technical capability becomes a product when a customer can reach value without a guided expedition. The task library translated computer vision into the language of a shift manager: queue, handoff, opening routine, service time. The dashboard translated detections into patterns. The alert translated a pattern into a chance to intervene.
Wobot’s $2.5 million pre-Series A, led by Sequoia India in August 2020, gave the company room to expand the platform, hire senior talent, and move into new markets. It followed a 2019 seed investment from Titan Capital. Chhabra said in 2021 that Wobot had built a presence in India, the Middle East, and Southeast Asia and had entered the United States. Drive-thru operations became a particular focus, with cameras measuring queues, flow, and order times.
The overlooked asset was not a new dashboard. It was the footage nobody had time to watch.
Partnerships helped the software travel. In 2022, restaurant operations platform Jolt announced an integration with Wobot. In 2024, Interface Systems launched a Wobot-powered offering for quick-service restaurant and retail chains. Interface would handle camera assessment, upgrades, design, deployment, and maintenance; Wobot supplied the analytics. Chhabra framed the partnership around a familiar pain point: proprietary hardware and closed systems make useful insight needlessly difficult to obtain.
IV. The physical worldAI leaves the demo and clocks in
By 2025, Chhabra was discussing retail AI and Wobot’s partnership with AMD around NRF Europe in Paris. A session with AMD, Wobot, and Supermicro explored multimodal AI in stores and quick-service restaurants - combining video with transactional and contextual data, and processing more of it close to where the work happens.
The shift toward edge computing suits Wobot’s original temperament. A store cannot wait for a philosophical answer. It needs a fast one. Is the line growing? Has the shelf been empty for twenty minutes? Did the order leave the window? The useful unit of intelligence is not a grand prediction about retail. It is a timely observation attached to a decision.
It also reveals why Chhabra’s business background belongs in this story. Computer vision can be discussed as model accuracy, inference, and infrastructure. An operator asks a different set of questions. Who receives the signal? Can it fit an existing workflow? Does it improve the next shift? Can the system scale across locations without multiplying complexity?
His public remarks return to repeatability. In 2016, while describing Wobot’s earlier incarnation, he argued that success can be repeated only by following a tested procedure. Years later, the technology is more sophisticated, but the concern is recognizable. Wobot observes process, distinguishes the routine from the exception, and makes the exception easier to see.
V. The founder’s useful obsessionFall for the question, not the interface
When Chhabra returned to Purdue for an entrepreneurship panel in 2019, he gave students a sentence with the economy of a good product brief: “Be in love with your problem and that is how you will find the right solution.” It is easy founder advice to admire and hard advice to practice. Products accumulate code, customers, decks, and pride. Changing one can feel like betraying all four.
Wobot’s history offers a more interesting reading. Loyalty to a problem may require disloyalty to its first solution. The audit workflow became video intelligence. The guided enterprise product became self-service SaaS. The cloud platform moved toward partnerships at the edge. The question remained: how can an organization understand whether work in the physical world is happening as intended?
There are tensions inside that question. Cameras in workplaces demand clear rules, careful access, and credible privacy controls. Wobot says its product limits compromise to data privacy and offers video management around defined business tasks. The practical test is whether observation stays bounded to a legitimate operational purpose. Intelligence is useful; indiscriminate attention is merely expensive suspicion.
At its best, the product does something modest and valuable. It gives a human the right few seconds instead of asking them to watch eight hours. It lets a manager compare locations without standing in all of them. It makes a buried operational pattern visible while there is still time to respond.
The camera keeps its old perch above the door or checkout. The room looks unchanged. But somewhere between the lens and the dashboard, footage stops being a vault and becomes a note. Adit Chhabra has spent a career closing that small, consequential gap between what a company says should happen and what happened on Tuesday at 2:14 p.m.