The $45M Industrial AI Revolution | Daniel Raj David, CEO of Detect Technologies on BP2B S2 Ep.2
Daniel Raj David (guest), Amrut (host)
In this episode of Best Place To Build, featuring Daniel Raj David and Amrut, The $45M Industrial AI Revolution | Daniel Raj David, CEO of Detect Technologies on BP2B S2 Ep.2 explores detect Technologies uses AI to prevent industrial accidents worldwide today Detect Technologies ingests camera feeds, sensor/time-series signals, and enterprise data to produce actionable safety and efficiency interventions for frontline workers and executives.
Detect Technologies uses AI to prevent industrial accidents worldwide today
Detect Technologies ingests camera feeds, sensor/time-series signals, and enterprise data to produce actionable safety and efficiency interventions for frontline workers and executives.
The product focus is continuous, scalable risk detection aligned to OSHA-style rules, including automated real-world interventions like stopping machines when people enter danger zones.
The company’s moat is real-world industrial data: Detect claims over 10,000 TB of labeled safety-related visual information collected through early customer partnerships.
Detect pivoted from hardware-plus-software (sensors, drones) to a hardware-agnostic AI SaaS model during the pandemic while retaining staff by retraining operational roles into data-centric roles.
Global expansion accelerated after initial international validation with Shell, emphasizing that technology can scale globally but delivery, support, and customer success must localize by region.
Key Takeaways
‘Actionable’ matters more than ‘AI’ in industrial settings.
Detect positions value as interventions that change outcomes (alarms, machine stoppage, management reporting), not just dashboards or detection, because safety teams cannot watch everything 24/7.
Get the full analysis with uListen AI
Industrial safety is a leading-indicator problem, not a post-incident reporting problem.
OSHA-style rules codify scenarios that precede injuries (line-of-fire, work at height, vehicle risks, PPE), and the product aim is to spot these continuously when humans are not present.
Get the full analysis with uListen AI
Real-world data collection is a defensible moat in applied AI.
Detect argues synthetic data was not viable when they started and remains inferior for full realism; partner deployments enabled them to build large, labeled datasets (stated as 10,000+ TB).
Get the full analysis with uListen AI
Hardware can be an on-ramp, but software scale often requires hardware-agnosticism.
They began with patented sensors and drone-based acquisition, but COVID constraints forced a pivot to run on existing customer infrastructure—accelerating deployment speed and international reach.
Get the full analysis with uListen AI
Edge vs cloud is a capability trade-off driven by rule complexity and bandwidth.
Some rules can run on-device for remote/offshore sites, but a full safety/compliance rulebook (“1000+ rules”) typically needs cloud or more powerful on-prem compute.
Get the full analysis with uListen AI
International growth in B2B is frequently reference-led and execution-sensitive.
After proving value with a credible global partner (Shell), adoption spread through industry networks; however, regional entities and strong customer success were required to meet higher delivery standards.
Get the full analysis with uListen AI
Deep-tech startups benefit disproportionately from a dense university ecosystem.
IIT Madras provided lab infrastructure (CNDE), mentorship (incubation cell, industry veterans), and the CFI student talent pipeline—enabling large student teams to prototype and iterate with industry.
Get the full analysis with uListen AI
Notable Quotes
“Our mission… is to make the industrial world a better place and a safer place through actionable artificial intelligence.”
— Daniel Raj David
“You need something that is looking out for these risks twenty-four seven, three sixty-five.”
— Daniel Raj David
“We’re not gonna let go of a single person… [we] trained them to be data scientists… pivot completely to a AI SaaS company.”
— Daniel Raj David
“India, for us, is the biggest and toughest test bed in the world.”
— Daniel Raj David
“We have technology, we have professors, we have industrial specialists, but our X factor is students who don’t know that something can’t be done.”
— Daniel Raj David
Questions Answered in This Episode
Detect mentions ‘1000+’ OSHA-style rules—what are the top 20 rules that drive 80% of prevented incidents, and how do you prioritize what to productize next?
Detect Technologies ingests camera feeds, sensor/time-series signals, and enterprise data to produce actionable safety and efficiency interventions for frontline workers and executives.
Get the full analysis with uListen AI
You cite ~6,000 daily global frontline fatalities and ~2M deaths/year—can you share the exact sources, definitions (workplace vs industrial-only), and how customers respond to these numbers?
The product focus is continuous, scalable risk detection aligned to OSHA-style rules, including automated real-world interventions like stopping machines when people enter danger zones.
Get the full analysis with uListen AI
For ‘intervention’ use cases (machine stop, alarms), how do you handle false positives/negatives and liability—who owns the decision loop: Detect, the site, or the OEM?
The company’s moat is real-world industrial data: Detect claims over 10,000 TB of labeled safety-related visual information collected through early customer partnerships.
Get the full analysis with uListen AI
What does your data pipeline look like end-to-end (collection, privacy, labeling, model updates), and how do you ensure models generalize across countries, PPE styles, and site layouts?
Detect pivoted from hardware-plus-software (sensors, drones) to a hardware-agnostic AI SaaS model during the pandemic while retaining staff by retraining operational roles into data-centric roles.
Get the full analysis with uListen AI
Can you break down a real ROI case study: what incident rate reduction or downtime reduction did a customer see, and over what time period?
Global expansion accelerated after initial international validation with Shell, emphasizing that technology can scale globally but delivery, support, and customer success must localize by region.
Get the full analysis with uListen AI
Transcript Preview
Our mission as an organization is to make the industrial world a better place and a safer place through actionable artificial intelligence. In this ecosystem, we have technology, we have professors, we have industrial specialists, but our X factor is students who don't know that something can't be done. There are many levers in the ecosystem that can fuel that initiative. [upbeat music]
Hi, this is Amrut. We are at IIT Madras, my alma mater, and India's top university for people who like to build. We are here to meet some builders, ask them: What are you building? What does it take to build? And what makes IIT Madras the best place to build? [upbeat music] Hello, and welcome to the Best Place to Build podcast. This is season two. We are recording at the Center for Innovation in IIT Madras. Today with me is Daniel from Detect Technologies. Daniel, welcome.
Oh, thank you. Thank you for having me, Amrut.
Detect is one of the hottest, uh, IIT Madras startups of the last decade. Um, you've raised more than $40 million, and you have operations across the world. Um, and you're a, a, like, quiet startup, not too much noise. We don't see a lot of press releases. Um, so maybe you can start by telling us, what does Detect do, and, uh, what is your presence worldwide?
Firstly, thank you for having me. Uh, second, it's always good to, to give an introduction of, uh, Detect, right? So, um, I would say today, we- the company has gone through several evolutions in the time that we've existed, but today, our, our mission as an organization is to make sure or make the industrial world a better place and a safer place through actionable artificial intelligence, right? Actionable being the key word. We do this through a various set of technologies and means, right? But the overall summary of what we do is we built a central AI layer that can take in information, which can be visual from your cameras, time series data from sensors, as well as simple data that's, that's scribbled on notes and put into ERPs, right? And be able to process that for large industrial plants, as well as smaller fabrication yards as well, and put actionable information right in front of the frontline so that they can act on, let's say, something that is a life-saving risk that's happening at the facility. Um, and even post the required statistics to management so that you understand what's the culture of your company, how do I improve the efficiency of my company, the safety of my company, at scale, um, at an enterprise level. So we touch people both at the frontline all the way to the CXOs and the executives, um, of these organizations through very verticalized AI interventions, uh, from our end.
Okay, hold on, hold on. That was a lot. Let me take it in.
Sure. Sure.
Um, so you said industrial AI, uh, i- in, uh, in for safety and efficiency?
Install uListen to search the full transcript and get AI-powered insights
Get Full TranscriptGet more from every podcast
AI summaries, searchable transcripts, and fact-checking. Free forever.
Add to Chrome