War is changing — and in a hostel room in Hyderabad, two engineering students were already preparing for it. What began as late-night experiments with high-speed racing drones soon turned into a far bigger realisation: modern conflict would belong to those who move faster, adapt quicker, and think differently. As they built and rebuilt machines with their own hands, a defining question emerged — if technology can take the first hit, why should a soldier have to? That question became Apollyon Dynamics — a young deep-tech startup now working with the Indian Army to develop high-speed kamikaze drones and a Mobile Drone Manufacturing Lab positioned closer to the frontline. They have partnered with Indian Army, BSF, RPF, BITS Pilani and many more. In conversation with CE, founders Jayant Khatri and Sourya Choudhury reflect on identifying critical gaps in India’s drone ecosystem, working with military and more.
Excerpts
What led to Apollyon Dynamics?
We were engineering students building racing drones in a hostel room when the Pahalgam attack and Operation Sindoor unfolded. It was then that we realised how quickly warfare was changing. Strategically, unmanned systems were becoming central to modern militaries, and India needed to accelerate, especially given hostile neighbours. From our own hands-on work, we knew fast, inexpensive, rebuildable drones shouldn’t be treated like manned aircraft. If something is cheap, agile, and expendable, it should face uncertainty first — not a human. The core problem we saw was speed — of programmes and platforms. Defence drone development cycles are slow, often lagging behind shifting battlefield needs. At the same time, many drones prioritise endurance over time-critical missions, where speed equals survivability. Through high-performance FPV drones and continuous engagement with forces, we focused on closing that gap: fast programmes, fast platforms, and real-time operational feedback shaping design.
Cold-messaging Army officers on LinkedIn isn’t a typical startup strategy — what gave you the confidence to take that leap?
It wasn’t really confidence… it was a necessity. We didn’t have access to traditional defence networks or established suppliers. As students, there was no formal pathway to present what we were building to the people who actually understood the problem. What we did have was a working system and a clear idea of the gap it addressed. We reached out directly because we believed that if the capability was genuinely useful, the right people would be willing to look at it, regardless of who built it or where it came from. We weren’t trying to sell a product; we were asking for feedback from practitioners.
When we were invited by a colonel to Chandigarh, it wasn’t a lab test or a college demo… we were standing in front of officers who would judge our work purely on performance. There were no slides or promises to fall back on. We demonstrated several drones, including ‘Ahuti’, and each one had to do exactly what we said it would. The real risk wasn’t embarrassment, it was credibility. If the systems underperformed, the conversation would have ended there. When the drones flew as intended: fast, stable, and consistent. The tone in the room changed. The questions stopped being ‘Does this work?’ and became ‘What variants can you build?’ and ‘How can this be modified for different missions?’ That shift was the moment we knew it had worked. We weren’t pitching anymore, we were collaborating on operational problems.
What makes ‘kamikaze’ drones different from regular commercial drones?
Commercial drones are designed to be reused. They’re built around stability, endurance, and protecting the platform. Our ‘kamikaze’ drones are designed for speed, precision, and agility, at the cost of almost everything else. There is no excess weight, no recovery-oriented design, and very little tolerance for inefficiency. The systems are intentionally unforgiving: every component exists to serve a single mission in a very short time window, with no expectation of return. You can think of a commercial drone as a general-purpose tool. Ours are purpose-built systems — optimised to move fast, stay hard to counter, and complete the mission even if they don’t survive it. When we started out, there was no generic platform. We built drones to custom specifications given by different units. Each requirement was slightly different — terrain, range, speed, payload, launch method — so we adapted designs case by case instead of forcing a single solution. That approach gave us a very practical understanding of how conditions actually vary on the ground, and how quickly requirements can change.
What operational gap in the Indian Army’s drone capabilities led to the creation of a mobile drone manufacturing lab?
The gap was adaptability at the frontline. Drones were being used in fast-changing conditions, but manufacturing and repair were still happening far away. If a drone was damaged, lost, or needed modification, units had to wait for spares, replacements, or approvals while the operational situation kept moving. We saw that requirements were changing faster than logistics could support them. The Mobile Drone Manufacturing Lab came from that mismatch. Instead of sending drones back to factories, we moved assembly, repair, and modification closer to the unit. That way, systems could be rebuilt, adapted, or replaced in hours instead of weeks. In short, the lab exists to remove distance between the battlefield and the ability to respond to it.
What kind of mission-specific customisation can soldiers achieve when assembling drones on the move?
When drones are assembled on the move, soldiers can customise them very directly for the mission they’re about to fly. That can mean adding or removing payload-dropping mechanisms, changing landing gear to suit rough or confined terrain, or mounting different sensor packages like day cameras, night cameras, or low-light systems depending on the situation. Because this is done close to the unit, there’s no need to settle for a generic configuration: the drone can be built around the mission instead of the mission being shaped around the drone.
How does 3D printing inside the Mobile Drone Lab help adapt drones to different terrains and combat scenarios?
3D printing inside the Mobile Drone Lab allows us to rapidly prototype and iterate new designs without depending on CNC machines or centralised manufacturing. It lets us quickly build and modify attachments, mounts, airframe components, and mission-specific parts on demand, making the drones adaptable to different terrains and combat scenarios. Instead of waiting for new parts to arrive, designs can be adjusted and tested almost immediately, giving units the flexibility to reconfigure systems for a wide range of missions as conditions change.
What challenges did you face in turning frontline soldiers into trained drone manufacturers and technicians?
The biggest challenge was time. Frontline soldiers already have demanding schedules and are primarily occupied with regular military training and operational responsibilities. That leaves very little room to pick up complex new technical skills or spend long hours learning detailed manufacturing processes. So the challenge was on our side. We had to design systems that were simple to assemble, easy to repair, and intuitive to use, even with minimal training time. That meant reducing steps, standardising components, and removing unnecessary complexity. Once the systems were built around the realities of a soldier’s schedule, training and adoption became far more effective.
In sensitive regions like Jammu and Kashmir, how does this system reduce logistical vulnerability and downtime?
In regions like Jammu and Kashmir, long and exposed supply chains are themselves a vulnerability. Moving drones, spares, or specialised equipment back and forth increases both delay and risk. By enabling assembly, repair, and modification closer to the unit, the Mobile Drone Lab reduces the need for frequent resupply movements. Damaged drones don’t have to be sent back, and replacements don’t have to wait on convoys or central warehouses. That means faster turnaround, fewer logistical movements, and significantly less downtime. Units stay operational for longer, with fewer dependencies...which is critical in sensitive and time-compressed environments.
How does the Mobile Drone Lab strengthen India’s Atmanirbhar Bharat vision beyond conventional defence manufacturing?
Atmanirbhar Bharat is often interpreted as manufacturing equipment within national borders. The Mobile Drone Lab pushes that idea further — it decentralises capability itself. Instead of relying only on large factories and long supply chains, the lab enables assembly, repair, and adaptation at the edge, closer to where systems are actually used. That reduces dependence on centralised production, imported spares, and rigid procurement cycles. More importantly, it shifts ownership of capability to the user. Soldiers aren’t just operators of a finished product, they become part of the production and adaptation loop. That creates a system that is resilient, responsive, and far harder to disrupt.
What’s the roadmap ahead?
Our primary focus is doubling down on speed with autonomy. That means moving toward jet-powered unmanned systems, designed for significantly higher performance than conventional platforms. We’re working on systems capable of high subsonic flight, operating in regimes where response time, interception windows, and survivability look very different from what propeller-driven drones allow. We’re being deliberate about not overextending into unnecessary complexity. The goal isn’t novelty.. it’s building fast, reliable systems that remain true to our core philosophy: compress timelines, reduce exposure, and stay ahead of how modern conflict is evolving.