Building SportlogiQ

Left to right: Jake Chadwick, Mehrsan Javan, Marc Cuban, Craig Buntin, Philippe Desaulniers

Written by Omar Zahr, CTO of TandemLaunch

In January, we celebrated the acquisition of SportlogiQ by Teamworks, the closing chapter to a decade-long story of a company that had to be torn apart by forces beyond its control before it could resolve into the data analytics powerhouse it came to be. At the center of that evolution? Two TandemLaunch EiRs who could not appear more different on the surface, but whose shared values and vision elevated sports analytics to the technological frontier.

As with most TandemLaunch venture stories, before there was a company, there was a PhD thesis.

Mehrsan Javan spent his doctoral years in the Electrical Engineering department at McGill University, working under Prof. Martin Levine on a very un-athletic problem: How do you teach a computer to identify unusual activity in a crowded environment? It was an endeavor to build more intelligent safety and security systems that could interpret human movement from standard camera feeds without specialized hardware.

Craig Buntin had arrived at McGill through a door that had never been opened before and hasn’t been opened since. A seven-time Canadian national figure skating champion and 2006 Olympian, Craig had finished high school and gone straight to the ice. When he became determined to pursue an MBA, McGill admitted him purely on the strength of his GMAT score — the first and only time the institution had made that exception. 

Craig broke the siloed engineering department by walking through it, literally door to door, asking what each research group was working on. An entrepreneurial instinct driving a search for a problem to solve.

“This guy is crazy and I am crazy, maybe there’s something we can do together.” — Mehrsan Javan

Craig and Mehrsan met at TandemLaunch in November 2013. A meeting of two minds, one that lived in algorithms and systems, and another that understood human athletic performance from the inside. While they didn’t share a technical vocabulary, they immediately connected through their values.

Their first shared idea, in fact, had nothing to do with sports. Both had independently been chasing autonomous vehicles — Mehrsan’s activity-recognition work was exactly the perception capability self-driving systems would need — but it was far too capital-intensive for two first-time founders with no automotive background. Craig’s athletic background offered the natural redirect.

Ice, ice, maybe…

Their first pitch to Helge and the TandemLaunch team was delivered the same morning it was finished - 2 am. An analytics platform for figure skating events. Now they just had to make it real.

The first SportlogiQ prototype was tested at an international figure skating competition in Colona using three GoPro cameras and a scrappy collection of cables to a laptop. The system worked, and TSN was able to feature distance and speed metrics for the competition. Rather than bask in success, however, the two founders were dogged by an undeniable observation. The figure skating market was too small. What they had hoped could be product-market fit got demoted to proof of concept for a much larger beast, Hockey.

Manual Intervention

Anyone that tries to sell technology for its own sake is going to be very disappointed. The central design constraint that governed SportlogiQ’s decisions was not how the technology worked, but how to translate its output into a language that subject matter experts on the other side would value. Coaches, general managers, and scouts spend their careers developing judgements that no algorithm can replicate. Craig knew this innately. He had been on the other side of this dynamic, an elite athlete who understood, viscerally, what it felt like to have someone tell you they could reduce you to numbers. SportlogiQ had to approach this differently.

“You can’t take the decision away from them. Their experience cannot be replicated by technology. You have to find what makes their lives easier and focus on that.” — Mehrsan

Teams said they wanted more data on players, but more data didn’t mean more measurements, it meant more understanding. They wanted to see, concretely and consistently, how their players were making decisions. They also needed to see it in a form they could trust - video. Every data point that SportlogiQ surfaced they anchored to footage. It wasn’t the most comprehensive data set, it was the most legible and actionable signal.

“You have to figure out what the customer wants regardless of what you can deliver today, then work backward to figure out how much you can automate over time.” - Craig

The operating consequence of this insight was less elegant than the insight itself. SportlogiQ’s data collection — the layer that turned raw footage into the structured signals teams could navigate — was, in the early years, deliberately human-in-the-loop. Investors at the time consistently advised against this. The conventional wisdom was that no serious deep-tech company should be building a large operations layer alongside its technology stack; that scaffolding made startups un-investable. Craig is candid in hindsight that they did exactly what they were told not to do, on the simple basis that the automation was not yet ready to carry the full load alone, and the customer was not going to wait. That sequencing — customer needs first, technical scope second — defined the company’s product architecture for a decade, and the automation caught up to the vision over the years that followed.

Their first paying customer told that story plainly. An NHL team that seemed interested in technology but had no budget for it. The deal they struck; charge almost nothing now, and full price next year. SportlogiQ’s first cheque was for $6,000. The real treasure, however, was what arrived alongside it — a physical hard drive shipped to TandemLaunch with video footage of the entire league for the previous season. Mehrsan still has that hard drive.

What none of them could have predicted was that their first two team customers would each go on to win the Stanley Cup within their first couple of years working with SportlogiQ. Craig is candid that those were still early days: SportlogiQ was learning its customer base in real time, refining the product alongside the teams adopting it. But the teams that engaged most seriously with analytics turned out to be the teams whose decision-making improved fastest. Those teams won. SportlogiQ’s reputation in the league started there.

SportlogiQ signed their first full contract in 2015. By the time SportlogiQ spun out of TandemLaunch in February of that year, it had already begun building the customer network that would eventually cover 97% of the NHL.

Cancelled by COVID

The first crack appeared before COVID did. For most of 2019, SportlogiQ had been working toward a deal that would have transformed the company structurally — a contract with the NHL itself that would have converted them from a vendor selling into 32 individual team customers into the league’s official data provider. One large customer instead of thirty-two small ones. The strategic case was straightforward: simpler revenue, deeper integration, more predictable scale. The unstrategic case, which they would only fully appreciate later, was that they were trading thirty-two customer relationships for a single one. The deal fell through in early 2020. Years later, Craig would name this kind of risk explicitly — customer concentration risk — as one of the non-negotiables he now refuses to take on at a new company. At the time, it was just a deeply painful loss.

Then COVID arrived.

When the pandemic broke out in January 2020, SportlogiQ had 130 people and was burning close to a million dollars a month aggressively expanding their market capture. They had just closed a round that promised to help them break into soccer and American football. They were, in Mehrsan’s words, “pumping on all cylinders.”

And then the NHL cancelled its season.

Sports analytics is a nice-to-have. When there are no games being played, there is nothing to analyse. Within weeks of the league’s announcement, contracts were paused or being renegotiated downward. SportlogiQ’s revenue had effectively gone to zero.

The bad timing did not stop there. Since October 2019, SportlogiQ had also been advancing acquisition conversations with a publicly traded sports data company that recognized SportlogiQ’s potential as the gold standard for hockey video analysis. The prospective acquirer’s executive team had been physically in SportlogiQ’s office in March 2020 for full due diligence when the cancellation came down. The deal collapsed the same week it would have closed.

It was, by any honest accounting, the worst possible moment.

Cultural Cohesion, Company Disassembly

The natural reflex of a leadership team in crisis is to contract — to hold information close, project certainty downward, and buy time by saying as little as possible. Craig and Mehrsan did the opposite. They gathered all 130 employees, put the actual runway numbers on screen, walked through the scenarios under which the company would and would not survive, and said something that founders almost never say out loud, “If this is too much risk for your family, make the decision that’s best for you, and we won’t hold it against you.”

Nobody left.

The transparency was not an emergency improvisation. Mehrsan and Craig’s operating philosophy had always been to treat openness as the default and discretion as the exception. “Anything the team deserves to know, we share.” The COVID all-hands worked because it was not a one-time act. It was a consistent practice being applied at its sharpest moment.

Survival demanded a fundamentally different company than the one that had walked into 2020. The team describes what came next as the construction of a second SportlogiQ. The first had been built to chase potential — burn cash, expand markets, demonstrate scale, raise the next round. The second had to operate without permission from anyone outside the building. Craig puts the cultural shift in starker terms, “We had to stop dreaming about changing the world. Everything became unit economics in the next three months.” They brought in Mitchell Wasserman with an unsentimental mandate — tighten the business, hold the cost caps, keep the trains running on time — and Craig is the first to admit that SportlogiQ survived primarily because of Mitchell.

The harder part was internal. Craig’s natural instinct, by his own admission, had always been to build a culture of, “Upend the industry, make it bigger and better.” That posture had served the company well for five years. It was now actively dangerous. Mehrsan, whose work had been about pushing the technical frontier and opening new product lines, had to redirect his energy into iterative improvement and operational cost reduction. In Craig’s words, “he was always such an inspiration to work with when he was working on those game-changing technologies. The shift was tough.”

The reframe held. Two years later, SportlogiQ’s revenue had grown more than 20%, and the company launched into soccer. The discipline of the second company had earned them the right to try again.

Out of a Job

The clearest measure of whether a founder has built an institution rather than a personal vehicle is whether the institution can survive their absence. Mehrsan started learning that lesson in the summer of 2022, when SportlogiQ’s VP of Engineering — someone who had joined the company two weeks before incorporation — left to take an analytics role with the Montreal Canadiens. The departure left Mehrsan with fourteen direct reports overnight, in the same season SportlogiQ was launching three new products.

By any reasonable management theory, fourteen direct reports is not a workable structure. Mehrsan describes the realization with the kind of retrospective gratitude that only comes after the fact: the only way to make it work was to empower people to be more autonomous, let them make decisions, let them make mistakes, and hold them accountable to what they did. He stopped being the technical authority that approved everything and became the manager whose job was to make sure everyone else could approve their own work. The team that emerged from that period was structured around distributed judgment rather than centralized control, and in Mehrsan’s framing that was the entire point. I made myself irrelevant. I see that as a great success.

Craig describes the same realization from his own vantage, with the lens of someone who has now done this twice. The period when he felt he was at his most effective was the stretch when SportlogiQ was between zero and twenty people — when he could work with each person directly and the company moved at the speed of his judgment. After that, organizational layers and reporting structures necessarily inserted themselves between him and the work. He does not romanticize the smaller stage; he is clear that the company needed to grow well past it. But he names it as the precondition for the company being able to operate eventually without him. “A 20-person team today can accomplish what a 120-person team could five years ago.” The implication is not subtle. The reasons to build small, distributed teams are accumulating, and the founders who learn early how to make themselves unnecessary are going to be the ones positioned for what is coming next.

A Company is Bought, Not Sold

By 2024, SportlogiQ had stabilized into the kind of business it had spent the previous four years convincing itself it could become. Earnings were stable. The hockey market was effectively dominated. The soccer expansion was pulling its weight. The pre-COVID question of whether the company was a real business had been answered. The question that replaced it was what to do with the answer.

The natural move, and the one most founders in that position make, is to find a buyer. SportlogiQ had been here before — that earlier acquisition process had collapsed in the worst possible way, and the next several years had produced multiple inbound conversations that never quite cleared the threshold. In 2024, the founders decided to run the process properly: engage an investment banker, structure a real outreach, and put the company in the market with intent.

The interesting choice was what they did alongside it. SportlogiQ did not just position itself as a target. It positioned itself as a consolidator — an M&A platform in its own right. In October 2023, the company acquired a smaller player in the sports analytics space to demonstrate that SportlogiQ could be the buyer in this category, not just the bought. That single acquisition, intentional in both timing and signal, opened private equity firms as a viable acquirer class. Suddenly the field of credible outcomes was much larger than it had been a quarter earlier.

The bet behind the move was a structural reading of the market. Sports analytics, despite its visibility, is fragmented to the point of inefficiency — two or three publicly traded companies, a long list of mid-sized private ones, and a much longer tail of small specialists. The market was maturing toward the point where consolidation was no longer a question of whether but of who and when. SportlogiQ chose not to wait passively for the wave. They positioned themselves to ride it from either side.

The process took a full year of work. There were offers that came painfully close, but it was Teamworks, who SportlogiQ had been in conversations with since spring 2024, that became their new home.

What made SportlogiQ desirable as an acquisition target, in the end, was something Craig himself had not predicted at the start of the company. He had walked in a decade earlier with a clear strategic theory about where defensibility would emerge, “I always thought the data collection part would get commoditized, and the insights on the data would always be the thing we did really well.” By 2024 the opposite had happened. SportlogiQ’s data collection layer remained, by his own assessment, industry-leading by a landslide — durable, refined, and genuinely difficult for competitors to replicate. The insights layer, on the other hand, had been opened up by the arrival of large language models, which made dashboards and downstream features substantially more accessible to anyone with a good underlying dataset. Whoever acquired SportlogiQ wasn’t really buying the analytics layer, they were buying the dataset itself, and the unduplicable infrastructure that produced it.

Teamworks — a Durham, North Carolina company that has been quietly assembling itself into the operating system for elite sports — had already been using SportlogiQ’s player tracking data to power its own predictive intelligence layer. The acquisition formalized a relationship that already existed in practice, and folded SportlogiQ’s iCE platform — by this point the standard for automated video analysis across the NHL — directly into Teamworks’ broader infrastructure.

For SportlogiQ, the outcome is the kind of exit that vindicates the decade of work that produced it. For Teamworks, it is the addition of the deepest video infrastructure asset in hockey to a platform that already touches more than six thousand five hundred elite sports organizations globally. For the broader ecosystem, it is one more demonstration that the maturation of a vertical-AI category looks less like a single winner and more like a series of intelligent recombinations.

A Team You’d Never Trade

Asked separately, in interviews months apart, what they were proudest of from a decade at SportlogiQ, Craig and Mehrsan answered the same way and in the same order. Not the technology. Not the contracts. The team.

Craig describes it the way someone describes an extended family — the kind of group where, if your basement flooded at three in the morning, every one of them would show up with buckets. Mehrsan describes it through the unusual fact that SportlogiQ had employees who stayed for ten years, which is not a thing that happens in technology startups, and especially not in startups that burned through a pandemic, lost an acquisition the same week it would have closed, and had to dismantle and rebuild their entire operating model along the way.

The two founders give different answers to the inverse question — what they would have changed. Craig’s answer is structural and pointed: the original beachhead was too small. Hockey was a defensible market and a credible early proof of the technology, but it was a TAM ceiling, not a TAM floor, and SportlogiQ spent a decade trying to expand out from under it.

“You spend ten years chasing down a TAM that wasn’t big enough, looking for opportunities for TAM expansion. It’s really painful.” - Craig

Mehrsan’s answer is shaped differently and reveals a different relationship to the work. He volunteered, when asked, that there was probably nothing he could have done that he would not also be willing to do entirely differently. The clarification followed immediately: as long as the outcome was the same, the path didn’t matter. Decisions were instrumental rather than sacred. The thing that mattered was where the company ended up — the team, the business, the exit — and the route to that destination was simply the route that had happened to produce it.

“If I went back, would I do things differently? Maybe I would do every single thing differently. But if I achieved the same outcome, I would be extremely happy. I am a fan of experimentation. But I would not trade this ending for anything” - Mehrsan