This leader is working to create a world in which passwords are passe  

What if your bank didn’t need to recognise your face or your PIN, just the way you hold your phone? JP Mishra, Founder CEO of Deep Algorithm, is betting on exactly that. Building AI-native cybersecurity from India, he is reimagining digital trust from the ground up, one behavioural signal at a time. 

With over 18 years of experience in artificial intelligence, you have witnessed the security landscape evolve dramatically. How exactly does tracking subtle human habits like finger pressure or typing speed outsmart a sophisticated hacker better than a traditional password? 

Passwords are like a master key to your mountain of private and confidential information. If someone steals that master key, they can access the information, know your life inside out and know it so well that they can create a duplicate ‘you’ executing the most notorious attacks with you as the mastermind of it. That’s because with a password and OTP hack, now they are fully inside your system. 

Now, if your security infrastructure is designed to not just generate OTP, but also track your behaviour with the device ranging from speed of your typing, the pressure of your finger on the screen, the slight tilt of your wrist when you hold your phone — these are things that are uniquely yours and they can’t be replicated. A hacker can steal your password in seconds, but they can’t steal the way your nervous system works. 

This uniqueness of a user’s behaviour makes behavioural AI led cybersecurity a powerful tool to deal with threat actors because your unique behaviour can’t be a victim of a data breach. Passwords live in databases and those can be hacked or phished. 

And here’s the part people often don’t expect; it’s completely invisible to you. No extra steps, no codes to enter, no friction. 

You use an agent versus agent approach to fight off modern cyber threats. What does it look like behind the scenes when your autonomous AI actively battles a malicious AI to identify vulnerabilities? 

Think of it as a red team i.e. Attacker Agents and blue team i.e. Defender Agents running simultaneously, at machine speed, 24/7. The red agents are trained to think like hackers, invading your public environment, tapping into existing vulnerabilities in the system, building an attack map just like a hacker would after they have understood the existing security infra and its gaps, it will all happen in milliseconds. The moment a vulnerability is detected by the red team, our Defender Agents get to task. They don’t just raise an alarm but assess the risk and deploy a security patch – all this happens in real time. This leads to the window of taking action coming down from days to seconds. Discovery of threat, assessment and remedial action, everything happens in real time and that is how modern day cybersecurity should look like. 

Our Defender Agents are built to do exactly the same thing back. It’s two grandmasters playing chess in real time, except the board is your entire public attack surface and the game never stops. The organizations that will stay secure in this era are the ones who stop bringing human-speed defenses to a machine-speed fight. 

Deep Algorithm uniquely offers both fraud prevention and recovery. How does your intelligence platform track down perpetrators and help recover stolen funds within seconds of a financial breach? 

Our ability to not just prevent but initial recovery is a result of a series of innovations done in designing our security products. 

Most fraud solutions are built purely for prevention; they try to stop the bad transaction from happening. But the reality is, attacks are evolving rapidly and some of them get through. The second a fraudulent transaction is detected, our platform automatically triggers an investigation workflow directly integrated with I4C’s NCRP platform. We’re not waiting for a bank employee to file a report or investigate the fraud manually; the intelligence is flowing into the national system in real time, the moment the breach occurs. 

On the recovery side, our agents immediately begin mapping the money trail – tracing where the funds moved, which accounts they touched, and flagging mule accounts before the money gets layered further into the financial system. Because we’re integrated with NCRP, law enforcement and the nodal officers at receiving banks get actionable intelligence within seconds, which is the critical window to freeze funds before they’re withdrawn or moved offshore. 

What makes this powerful is the combination of behavioral intelligence and financial network analysis working together. We know how the fraudster behaved to get in, and we know where the money went after. That end-to-end visibility from the moment of breach to fund recovery is something no purely preventive platform can offer. In financial fraud, speed isn’t just an advantage. It’s the difference between recovery and permanent loss. 

You are building globally relevant deep tech right here in India while pursuing your doctorate at IIT Mandi. As financial scams continue to rise, what is your ultimate vision for creating absolute digital trust across our banking systems? 

India is at the forefront of digital financial inclusion, but with scale comes systemic risk. My vision is to move the industry toward “absolute digital trust” where fraud is not just detected, but fundamentally deterred. 

This requires three shifts: 

• From identity to intent: verifying not just who you are, but whether your behavior aligns with legitimate intent 

• From reactive to proactive security: stopping fraud before execution, not after loss 

• From siloed defense to collaborative intelligence: banks, fintechs, and regulators sharing real-time threat intelligence 

Building this from India gives us a unique advantage: we are solving for scale, diversity, and adversarial environments simultaneously. 

The end goal is simple but ambitious: an ecosystem where trust is embedded by design, and fraud becomes economically unviable.