SECOMIND: Artificial Intelligence as a Service

Ajay Malik, CEO

Despite being capable of hosting humongous amounts of data, the human mind can function with very little data. Every day, humans improvise whenever they encounter new data and get on with life. Time series analysis and incremental learning help humans to live daily life. But when it comes to AI, a common understanding is that an organization needs to have a plethora of data to gain an advantage of leveraging AI. Challenging this concept is SECOMIND through its AI as a service to help small and medium businesses (SMBs) get started with their AI solution. SMBs lack the required budget and resources to solve a particular problem using AI, and that’s where SECOMIND comes into the picture.

SECOMIND helps SMBs to solve their problem without spending too much on implementing or developing full-scale AI models and hiring resources to run them. “The clients can reap the benefits of AI within weeks of implementation. They can start from any machine or device and make it an intelligent cloud-managed solution,” says Ajay Malik, CEO, SECOMIND.
Former Head of Architecture/Engineering of Worldwide Corporate Network at Google, Ajay started SECOMIND in 2021 and now has over 130 employees, and R&D centers in 5 countries.

The company has also developed its own IoT AI cloud platform called CLEA which hosts ready-made solutions that can be deployed in a client's environment in a matter of weeks.

SECOMIND specializes in self-supervised, incremental, and reinforcement learning models. Each of the solutions can mimic the human brain, where there is continuous prediction and analysis of the available data at any moment. It can correlate distinct datasets and perform descriptive, diagnostic, prescriptive, or predictive analysis for better sales, inventory or demand forecast, or inventory management as per the client’s requirements. The team at SECOMIND is adept at developing these solutions even with a very small set of data. And as the data grows, the solution keeps learning, recognizing, and detecting newer problems and enhances the AI model for better prediction. In addition, the models can be further improved by incorporating feedback on the results. SECOMIND builds AI that can run in the cloud, on the edge or even in FPGAs.

SECOMIND has deployed AI solutions for Vending Machines, EV Chargers, Drone Search & Rescue, Drone Detection & Tracking, Sales forecasts, Supply Chain demand management, Surveillance, Fitness Support, Safe Driving, Cybersecurity, and more.
One such organization was a smart cooler manufacturer that wished to identify unit models sold and update its inventory accordingly since they were constantly building newer models. RFID technology was beneficial to an extent but accurate inventory management required a deft product recognition solution. SECOMIND was called in to help. The company implemented an incremental learning model using image recognition which then identified different types of models. Subsequently, as the newer models were introduced, it was able to add them to the inventory list, creating a well-functioning dynamic system.

The clients can reap the benefits of AI within weeks of implementation

To create a solution that works like a human brain, SECOMIND not just invests in Data Scientists and AI/ ML engineers but also employs neuroscientists who help in adapting brain-like functions to their offering. The knowledge comes in handy in creating new AI models. And while SECOMIND develops chatbots, it also understands the need for human interaction in troubleshooting and runs a live chat center and 24x7 AI Ops for its clients. For the future, Malik is hopeful that their R&D in understanding the human brain and AI/ML expertise will result in chatbots that can soon do real-time troubleshooting for the clients.