He Raised $74 Million To Bring Real-Time Intelligence to Warehouses With AI-Powered Vision
AI and robotics are having a moment. But very few founders are actually building at the intersection of deep tech, real-world deployment, and scalable business models. Gather AI CEO & co-founder Sankalp Arora, PhD, is one of them.
From growing up in Delhi with a childhood obsession with robots to building one of the most advanced warehouse-intelligence platforms in the world, Sankalp’s journey is a masterclass in conviction, technical depth, and execution under pressure.
This is the story of how curiosity turned into company-building—and how robotics is quietly reshaping global supply chains. In a riveting conversation on the Dealmakers Podcast, Sankalp talks about getting into Carnegie Mellon without spending a dime and receiving awards from media outlets.
A Childhood Wired for Robotics
Sankalp didn’t “discover” robotics later in life. He was wired for it. Growing up in Delhi in a joint family, surrounded by cousins, cricket, soccer, and academics, he was always drawn to one thing: building machines that move and think.
“Since I was 5 years old and can remember, I just wanted to make things that move, that can think,” Sankalp recalls. For many in India, engineering is a default path. For him, it was something deeper—it was instinct, “a very natural proclivity, just a neural network being wired like that,” he remarks.
After doing his undergrad at Delhi College of Engineering, Sankalp continued building small robots with limited resources. But he knew that to truly push boundaries, he needed to be in the best space possible. That meant one place--Carnegie Mellon University (CMU).
During his undergrad studies, Sankalp had worked on India’s first truly autonomous tank. One of the people he had worked with referred him to a professor at CMU.
Betting Everything on Carnegie Mellon
Carnegie Mellon University is widely regarded as the epicenter of robotics innovation. It would have a pivotal role in Sankalp’s life. He would go on to get every degree possible from the university, which is in itself a spectacular feat.
However, getting in is hard. Affording it is even harder. Sankalp didn’t have the financial means—but he had something more valuable: proximity to opportunity. Through a connection, he reached out to Professor Sanjiv Singh and struck an unconventional deal.
If he proved himself within three months, Professor Singh would give Sankalp a staff position, which would fund his education. Sankalp’s father had to take a loan just to pay for the flight ticket to the US. But three months later, the bet paid off.
Sankalp earned the position, and his master’s and PhD were fully funded through the lab and the Department of Defense. While at CMU, he worked on groundbreaking projects—including the world’s first fully autonomous helicopter. But one idea stood out.
The Core Insight: Make Robots Curious
During his PhD, Sankalp focused on a fundamental question: How do you make robots curious in physical environments?
This wasn’t just academic; it became the foundation of what would later become Gather AI, which uses autonomous devices and AI-equipped cameras to monitor operations across warehouses.
Toward the end of his PhD, Sankalp realized that the most direct way to have a positive impact on the community through his inventions was entrepreneurship. Founding companies would actually enable him to meet the people whose lives his inventions would benefit.
By this time, Sankalp had published a few papers that focused on making robots autonomous using only cameras. These papers had attracted interest from some of the world's largest autonomy companies to license the technology. The timing was perfect to test the impact in real-life spaces.
Considering this was his first job ever, Sankalp had a lot to learn. Although he had a deep knowledge of the robot autonomy and the curious robots domain and space, he knew very little about the outside world. Thus, he signed up for VentureBridge, an innovation fellow and CMU incubation program.
As part of this program, Sankalp received funding from the Department of Defense for customer discovery.
Finding the Right Problem: 175 Conversations Later
Great founders don’t just build technology—they match it to urgent problems. Using the funding, Sankalp and his team conducted 175 customer discovery interviews. As he recalls, they had the “hammer” (cutting-edge robotics + AI). Now, they needed the right “nail.”
Sankalp and his team found it in supply chains and supply chain visibility. They realized that if they could make curious robots for operations within the four walls of a warehouse, it would solve an urgent need. They could address a large enough market and have a global impact.
This is where their core thesis of using AI to gather data comes in. As Sankalp reveals, the fascinating part was that other companies, including YC-funded ones, had tried this before and failed to materialize the technology as intended.
When Investors Said “This Isn’t Possible”
The early fundraising story is a classic deep-tech paradox. The tech was groundbreaking, which made investors skeptical. The default response Sankalp heard: “This doesn’t work.” So he did something unconventional. He brought the product to the investors.
Sankalp built mock warehouses in Salesforce Tower offices and flew drones during pitch meetings to demonstrate the technology in real time. At the same time, he secured early customer validation: “If you build it, I’ll buy it,” they said.
These customers had full faith in the tech stack and the tech ability. That combination—live proof + customer pull—unlocked the first funding. This was sometime in 2019.
What Gather AI Actually Does
At its core, Gather AI turns ordinary cameras and hardware into intelligent data collectors. As Sankalp explains, they can use the most basic hardware available at Best Buy, whether for drones or forklift-mounted cameras. Next, they digitize workflows and inventory data in warehouses using cameras.
By digitizing those workflows, Sankalp and his team offer not just real-time visibility but also enable warehouses to ship more goods on time and fulfill orders in full. At the same time, they are reducing the amount of labor they spend on shipping each and every item.
From the customer’s perspective, all they need to do is click on a product on an e-commerce site and have it magically appear at their doorstep. Gather AI provides the backend technology that makes the magic happen—as promised by the e-commerce vendor.
The First Signal of Product-Market Fit
Sankalp realized that Gather AI had built the world’s first such technology. No other team or tech stack can take a camera and transform it into an effective data-gathering tool using just software. He had immense faith in his team’s tech capabilities.
Before scaling, there was a clear moment of validation. After launching a basic website—with zero outbound effort—Gather AI generated $5M in pipeline within a month. That told Sankalp two things. One is product market fit, and the other is the core technology.
People were reaching out to Gather AI, affirming they needed a solution to their problem. It still took three years to fully develop the technology. But when it was ready, customers didn’t just adopt it—they expanded rapidly across multiple facilities and have continued to grow their footprint.
As Sankalp points out, landing customers was relatively easy because the problem was easy to solve. Next, they started scaling the solution rapidly—both were distinct checkpoints in the product-market footprint.
Raising $74M—and the Power of the Right Investors
To date, Gather AI has raised $74M across multiple rounds. Sankalp reveals that they first raised a pre-seed round of ~$3M, and then, a seed round worth $10M led by Expa and Xplorer Capital. He considers himself fortunate to receive support from top mentors and credible people.
For instance, Expa is a startup studio and venture fund founded by Garrett Camp, who is known as the co-founder and first CEO of Uber. Xplorer Capital was founded by serial entrepreneurs who have sold multiple companies to Amazon. Kiva Robotics is one of them.
Gather AI’s further g rowth rounds included $10M from Tribeca Venture Partners, $17M from Bain Capital Ventures, and $40M Series B from Smith Point Capital (cofounded by Keith Block, the ex-CEO of Salesforce).
These investors weren’t just funding the company—they were reinforcing conviction.
Storytelling is everything that Sankalp was able to master. The key is capturing the essence of what you are doing in 15 to 20 slides. For a winning deck, take a look at the pitch deck template created by Peter Thiel, Silicon Valley legend (<a href=" target="blank" rel="noopener">see it here</a>), where the most critical slides are highlighted.
Remember to unlock the pitch deck template that founders worldwide are using to raise millions below.
The Near-Death Moment: COVID
Right before COVID hit, Gather AI was about to sign a major multi-facility deal with one of the top global retailers. Six months of negotiations had resulted in a verbal confirmation on Friday that the final signatures were expected on Monday.
Sankalp and his team were looking forward to the significant revenue they were expecting to earn from the deal. That just disappeared. Further, all the warehouses in the US shut down to external vendors and partners as they figured out how to operate on their own.
For the next six months, no one was allowed to be in any of the warehouses. As Sankalp sees it, the challenge of surviving centered on having no new leads for those six months. But talking to people in warehouses made him realize the trust and conviction Gather AI had earned.
Investors and current employees continued to invest in Gather AI. They were confident that when things opened—and they would eventually open—they would find enough business, which did happen because Gather was solving a relevant problem.
When warehouses reopened, demand surged—and Gather AI was ready. Sankalp talks about how their conviction was really tested, as all they could do was double down on product and engineering despite having no sales at all.
The Bigger Vision: Fixing the Global Supply Chain
As Sankalp points out, most consumers may not understand the real impact or experience much change other than receiving deliveries on time. Though instead of same-day delivery, they can now expect deliveries within a couple of hours.
But for the people working in the warehouses, efficiency means not having to run around the warehouse looking for lost items. It means they can go home much earlier to their kids.
Today, inefficiencies in supply chains cost companies billions. Retailers often hold 5% to 15% of annual revenue in inventory. For a $5B company, that’s up to $750M sitting idle—around $0.5B worth of extra inventory in warehouses, depending on how logistics are operating.
Gather AI’s vision is to eliminate this inefficiency. In a fully realized future, Sankalp sees this kind of technology helping to protect supply chain surprises arising from geopolitical factors, such as the pandemic. Supply chains become resilient to disruptions, and warehouses operate with real-time intelligence.
Who Wins: AI vs Robotics?
Sankalp has a nuanced view of the AI landscape. His eight years as an entrepreneur have taught him that the people who will win are the people who deliver value and own the workflow.
In most desk jobs, workflows are owned by large language models (LLMs), not physical AI. Any white-collar work is mostly catered to and owned by large language model-driven workflows. Or it can be something in the digital domain, unless a better architecture than the current transformer emerges.
In Sankalp’s perspective, physical AI is still quite young because the core principles stay the same. As a result, in white-collar work, LLMs can be closely monitored and supervised. Any ramifications can be dealt with, and errors can be quickly corrected.
But in physical environments, the tolerance to mistakes is much lower. The consequences are much more serious in cases involving robot-related injuries or the endangerment of human life.
As a result, a longer tail is needed—much like autonomous cars, which are the first large-scale physical AI humans are fielding.
But eventually, most of the actions we perform manually or move around will be taken over by robots. Sankalp anticipates that these tasks will be taken over by humanoid robots.
Specific solutions that optimize the efficiency of particular workflows industries need will likely emerge. Physical AI could be a humanoid butler at home, where, beyond efficiency, people care about flexibility. Even then, large language models will always be needed.
As Sankalp underscores, large language models can win on their own simply because they can take over white-collar work and automation. Winning without large language models will be challenging.
Building a Team of 75 Around One Principle
At Gather AI, hiring isn’t just about technical excellence. Two traits matter more than hiring the best in their domain in the world—people who are the best in building physical AI that actually works in the real world and not just in a compute cluster.
Customer Compassion: People who deeply understand and care about the problem and how it can help someone’s life. As a result, Gather AI has attracted employees who have experienced this problem themselves or in some form at work. For instance, people from large supply chain companies like Amazon, Walmart, and Uber, as well as other deep tech companies. Yet another example is people from autonomous car companies with an ecosystem in Pittsburgh. · High Agency + Curiosity: People who take ownership and explore beyond their domain. Interviews at Gather AI are designed to filter and identify such talent. That’s how Sankalp has recruited the 75+ people at Gather AI. They are a group of people who are really driven to solve someone's problem and take ownership of it. They are curious about building because building a product like Gather AI takes full-stack machine learning, autonomy, mobile development, customer success, solutions, and sales—all working together across the board.
Lessons for Founders
If the mission is right, people will align with you and join you in the journey: Early doubt about attracting talent is natural—but conviction compounds. As Sankalp has worked most of the time independently, he believes in owning his thesis, even if a large team contributes to it. · Hire carefully early on: Early hires shape culture and trajectory. It was hard for Sankalp to predict how people would join his journey and if they would have the same dedication and curiosity to solve customer problems. He believes in carefully vetting hires and getting to know them up front. · Deploy capital aggressively: “Investor money is not meant to sit in the bank.” Being too conservative early on costs momentum.
Final Thought: Curiosity as a Competitive Advantage
Sankalp’s journey—from a kid in Delhi dreaming about robots to building a category-defining company—comes down to one trait: curiosity, and not just in machines. But in order to understand problems deeply, challenge assumptions, and push boundaries in the real world.
In a world flooded with AI hype, the companies that win won’t just build models; they’ll build systems that work in the real world. And that’s exactly what Gather AI is doing.
Curiosity compounds into company-building when paired with the selection of real-world problems. · Deep tech only works when it translates into clear, measurable customer value. · 175 customer conversations beat assumptions every time in finding product-market fit. · Proof in the real world converts skeptics faster than any pitch deck ever will. · Strong pipeline pull is the clearest early signal that a problem is worth solving. · Conviction during zero-revenue periods is what separates survivors from casualties. · Winners in AI will be those who own workflows, not just build technology.
Original Version
Alejandro Cremades: All right, hello everyone and welcome to the DealMaker Show. So today we have ah an amazing founder, you know, founder in robotics. We're going to be talking a lot about the building, the scaling, the financing and robotics. My God, there's there's quite a lot of...
Alejandro Cremades: of AI and robotics talk going along these days. So quite timely, the episode that we got. Also, they've been receiving awards left and right from from very recognized a media outlets. So you're going to be really being part and listening to what I think is going to be a very inspiring conversation, you know, where we're going to be covering two how he got into Carnegie Mellon without spending a dime, really surviving COVID, which was a kind of crazy for them, or or coming from a background of really deep tech, but allowing themselves to bring something to market that is changing people's lives. So again, really incredible conversation ahead of us. So without further ado, let's welcome our guest today, Sankal Arora. Welcome to the show.
Alejandro Cremades: So originally born in Delhi and raised there too. Give us a walk through memory lane. How was life growing up for you?
Sankalp Arora: A lot of fun, I would say. all I grew up in a joint family with all the cousins living around in the same apartment complex. and had a lot of fun playing cricket, soccer and studying a lot.
Sankalp Arora: I was a nerd growing up, always wanted to make robots. My cousins tell me that since I was five years old, I wanted to make robots. I don't think I knew what robots were at five years old.
Sankalp Arora: But fortunate enough to have followed that path after after doing my undergrad at Delhi College of Engineering. engineering
Alejandro Cremades: That's incredible. so So tell us, too, about the engineering and solving problems. You know i know that obviously there is quite the pressure in India, you know, for either becoming a doctor or engineer. But in your case, you know, engineering, why?
Sankalp Arora: Oh, it's very hard for me to answer why, because since I can remember, i just wanted to make things that move, that can think. And so it's a very natural proclivity, just a neural network being wired like that.
Alejandro Cremades: Now, coming to the U.S., um obviously Carnegie Mellon had a quite a pivotal role in your life and career. um i mean, you literally got every degree that anyone can think of you know from the university, which is a spectacular. but But tell us, coming to the U.S., Carnegie Mellon, what was that journey like? Because that was incredible for you.
Sankalp Arora: while lama in my undergrad, I'd been making little robots that whatever little funding in the undergrad would would allow or enable. But knew sort of the best school in the world to learn robotics is Carnegie Mellon.
Sankalp Arora: And one of the people that I worked with during my undergrad, we worked on India's first fully autonomous tank. He referred me to a professor here at CMU.
Sankalp Arora: I didn't quite, my me and my family didn't quite have the means to afford an education at Carnegie Mellon. ah But there was an understanding with my professor, Sanjeev Singh, who said that if you come here and I like your work, I'll offer you a staff position.
Sankalp Arora: And that will in turn lead to your master's being paid. So my dad took a loan to fund my flight tickets. And I was here.
Sankalp Arora: Fortunately enough, in those three months, Sanjeev really liked my work and offered me the staff position. And I did my master's and PhD covered both by Sanjeev's lab and Department of Defense ah from then there on forward.
Sankalp Arora: The most exciting part of being at CMU is you get to work on stuff that is just unimaginable. So I got to work on the world's first fully autonomous helicopter, which is like a childhood dream come true as a far as projects go.
Sankalp Arora: And then focused my thesis on how to make robots curious in physical spaces. And an expression of that is in our company, Gather AI, which is gathering data using AI through curious robots.
Alejandro Cremades: So then talk to us too about Gather AI and how Gather AI comes together because this was obviously coming out of some of the work that you were doing at the Carnegie Mellon and and also the way that you guys spin it out. um So talk to us about how the whole origination of the idea and and how you guys went about incubating it and then bringing it to life.
Sankalp Arora: So i I think towards the end of my PhD, I realized that a way to the most direct way to have a positive impact through my inventions on someone's life is entrepreneurship, where of I can actually meet people whose life my invention seems maybe a little bit in in a positive fashion.
Sankalp Arora: And we had just published a few papers around how to make robots autonomous just using cameras and had gotten interest from some of the largest autonomy companies in the world to license out the type technology.
Sankalp Arora: So it seemed like the perfect time to also see if we can have that direct impact on the world. So we joined not having worked before. This is my first job ever.
Sankalp Arora: So it was quite a bit to learn. and And we knew deeply about our domain and space of of robot autonomy and curious robots. But very little of the world outside of that.
Sankalp Arora: There's a CMU incubation program called Venture Bridge, an innovations fellow. As a part of that, we and then Department of Defense funded us for customer discovery.
Sankalp Arora: We did about 175 customer discovery interviews. from we had the hammer to find the right nail, which was supply chain and supply chain visibility, and decided that if we can make curious robots around what's happening within the four walls of a warehouse, it's an urgent need that we can address of a large enough market that we can have a global impact.
Sankalp Arora: So that's where the core thesis of gathering data using AI comes from.
Sankalp Arora: The fascinating part was There were other companies, including YC funded ones, that had that had tried this before and failed to materialize the technology as it should.
Sankalp Arora: So whenever we went out to market to raise, most of the people said that this is not possible because others have tried it. And and it it just doesn't quite work.
Sankalp Arora: And for us, really, the turning point came when during doing those investor interviews, we were also able to get a couple of customers on board to say, I've seen this in my warehouse because we could we had full faith in our tech stack and our tech ability.
Sankalp Arora: And if they build it, I'll buy it. Plus, in SF, ah we actually ended up you know right near Salesforce towers. We went to a couple of investor offices.
Sankalp Arora: We actually ended up flying drones and setting up a mock warehouse in little investor offices to show them the stuff working. And that's what got us our first funding game. That was back in early 2019, right after my PhD.
Alejandro Cremades: So I guess i guess let's rewind back, you know just so that the people get it.
Alejandro Cremades: What ended up being or becoming Gather AI and how do you guys make money?
Sankalp Arora: So we digitize workflows and inventory data within warehouse through cameras that you can buy out of Best Buy, whether it be drones or cameras on forklifts.
Sankalp Arora: By digitizing those workflows, we then offer not just real-time visibility, but the ability for these warehouses to ship more goods on time and in full, while reducing the amount of labor they spent on shipping each and every item they ship out.
Sankalp Arora: A simpler way of saying it is, for most of the people who click and the thing appears on the door magically,
Sankalp Arora: we enable that for larger percentage of that time, that thing appears on time as promised to you by that retailer or by that e-commendor.
Alejandro Cremades: Now, talk to us then about the way that you guys finance the operation. And I guess even before that, what would you say was the first early sign that really got you thinking that they had this had a lot of potential and that it was going to work?
Sankalp Arora: There are two ways of looking at work. One is product market fit, and other is the core technology. Because what we were building and have built is a world's first, where no other stack or team in the world can can just take a camera and turn it into an effective data gatherer with just software. ah So this we had we had quite a bit of faith in our in our tech capabilities. The product market fit, I think, really showed itself When we raised in in early 2019, mid 2019, we launched our very basic website with no expert advertisement at all.
Alejandro Cremades: So that's fantastic. That's fantastic. Obviously, you see so many lessons too that you've learned along the way, whether it's from building the company or building the team, especially because you've been pushing this for quite some time now, but...
Alejandro Cremades: Imagine if I was able to bring you back in time. Okay, so I bring you back to the days of Carnegie Mellon. And let's say while you're getting your PhD and the idea of Gather AI just pops across, let's say I'm able to bring you back to that moment in time.
Alejandro Cremades: And to that moment where you are also able to tell that younger self and give that younger self one piece of advice before launching the business. What would that be and why, given what you know now?
Sankalp Arora: I think the first and most important thing would be just the confidence that if the mission is right, people will align with you and join you in the journey.
Sankalp Arora: Since I worked just by myself most of the time, and even as a PhD, your thesis has to be your own, even if a large team contributes toward it, it was very hard for me to picture how people would come along on this journey and how people would feel other people's pain.
Sankalp Arora: Second is, make sure that early hires are really well vetted. Because I think we made some initial hires that I feel like we could have done better on.
Sankalp Arora: So just knowing that upfront. And third is upfront knowing that the money that investors invest is not for keeping in the bank but to be deployed.
Sankalp Arora: I think in the early days, I was too stingy with deploying that money. And that cost the company a bit of momentum.
Alejandro Cremades: I love that. So for the people that are listening that would love to reach out and say hi, what is the best way for them to do so?
Sankalp Arora: Often it's LinkedIn, actually. It's very searchable, right? So if you look at Sankalp Arora and Gather AI, or you can reach us through our website. Just say hello at gather.ai, and those emails come to me and we can connect.
Alejandro Cremades: Amazing. Well, Sankalp, thank you. It's been an amazing, amazing time here that you've given us. It's really an honor that you had the time here to dedicate to us, and I really appreciate you and all this incredible wisdom that you've provided us. So thank you so much for being on the DealMaker Show today.
Sankalp Arora: Thank you for having me, Alejandro, and letting me share whatever little I've learned.
