In 2010 years ago Tarun Wig and Abhishek Sharma started on a journey to develop technology which could bring changes into how organisations handle their data. While most of the organisations in the country were focused on providing services, they wanted to focus on developing state of the art products, indigenously in the country.
Out of this thought, Innefu, an AI-driven company which develops cutting edge technology to carry out predictive intelligence and cybersecurity solutions, was born. The company which started with a team of ten people working out of the living room of a house is today more than 100 people strong with clients spread across three different continents. Innefu today serves four out of top 10 corporates in the country apart from serving some of the largest and most critical Law Enforcement and Intelligence organisations in South East Asia.
Analytics India Magazine talked to Wig to know more about the startup and its security solutions using data, in India and abroad.
Subscribe to our Newsletter
Join our editors every weekday evening as they steer you through the most significant news of the day, introduce you to fresh perspectives, and provide unexpected moments of joy
The tech team at Innefu Labs comprises of data scientists, NLP experts, computer vision experts, application developers and data mining engineers. Predominantly everyone in the team holds an MCA or an MTech or at the very least a BTech with substantive work experience.
The startup uses its own analytical tools for carrying out an explorative investigation or predictive intelligence.
Innefu integrates multiple data sources in their predictive intelligence solutions::
- Data from social media platforms including Twitter, Facebook etc.
- Data from Newspapers
- Data from websites, forums etc.
- Internal Datasets available with the customer in unstructured forms
- Data from CCTV Cameras and images
When asked about what every day of Innefu’s tech team look like, Wig jokingly says, “I’ll use three terms — coffee, lunch breaks, and snacking in the office.” The tech team is divided into three parts:
1.R&D for developing new AI models for Vision and Predictive Intelligence
2.New feature developments in the product
3.Deployment in existing customers
The company uses low-level C for video and image analytics. Most of the other ML models are in Python. Some of their models include RNN, Logistic Regression, K-means and SVN.
Innefu Labs Solutions
Innefu has the same underlying AI framework that we use across three different products of:
1.Prophecy — Big Data Analytics: Prophecy is an AI-driven framework specifically tailored for Predictive Intelligence and Predictive Policing. The solution is deployed across multiple Law Enforcement and Intelligence agencies.
Prophecy ingests data from multiple sources including raw documents, forms, PDF’s etc apart from other relational Databases in one Big Data lake. The solution augments the dataset with a lot of data extracted from Open Source databases and extracts intelligence from raw data. While it was initially only available for LEA’s, we have recently released Prophecy versions for Retail analytics, Insurance Fraud and Cyber Security.
2.Vision — Video and Image Analytics: Vision is a Video and Image analytics solution which takes data from live CCTV cameras, apart from images and recorded videos and automatically provides alerts to the analyst on the identification of a known criminal, identification of weapons, unclaimed bags etc. The solution has multiple models for security and surveillance including detection of an unauthorised individual, trespassing in a secure area, fence breach etc
3.AuthShield — Authentication Security: AuthShield is a unified Authentication platform which integrated Facial and Thumbprint Biometrics with standard 2FA solutions including Push notifications, OTP’s, SMS and Email OTP’s etc
Some of the biggest clients of the startup include DRDO, paramilitary forces including CRPF, BSF, Assam Rifles etc., the State police departments including Delhi, Mumbai, Punjab, multiple wings of Indian Army and corporates such as Mahindra and Mahindra Group and Innodata.
One of the biggest challenges the startup faced in the beginning was that the startup ecosystem at that time was almost non-existent. There was a huge mind block in trusting a new product out of an innovative startup viz a viz an old technology from a reputed brand.
While India was gaining a reputation in providing services, there was no inclination to evaluate and ultimately buy Indian products. Being one of the first Indian companies to venture into product development the founders initially struggled to break into this mindset. They had to ensure that they worked hard at every stage because they knew that even a single failure would be a disaster for them with no looking back. “Luckily for us, we carried that thought process as we went ahead and we haven’t failed in any of our targets,” said Wig.
Innefu Labs has seen some amazing successes in their journey. Following are some of them:
- Deployed a statewide data monitoring system with 100% delivery in 2012.
- Played an instrumental role with a government organisation in tracking Ghostnet attacks in India. They were able to identify 496 organisations which had been compromised using 122 C&C Centers.
- Cracked a hacking cell operating from three different countries and hacking into Indian users.
- Implemented globally the biggest authentication security framework for 2 million users.
- Is the first Indian company to implement Predictive Intelligence framework across 5 different LEA and Financial fraud investigation agencies.
Their biggest achievement was when Delhi Police used their video and image analytics solutions to identify 3,000 missing children in four days. It is a case study which was quoted all over the world as a first actual use case of AI for social welfare.
While talking about the difference in deploying analytics solutions in India from those in other countries, the Founder said that the solutions remain the same, however, the sources of data differ drastically. The data sources outside the country are much more structured and streamlined, making the job easier.
Talking about the threats that technologies like Analytics and ML can cause to the security, Wig said, “The data volume is increasing every day. It is imperative to analyse all the data using multiple AI models to ensure we are able to identify anomalies and patterns not evident to the naked eye.”
Innefu Labs does not have any Indian competitor but only international competitors. Palantir, Verint Systems, Cobwebs Technologies are some companies working in this space and are its competitors.
Innefu goals for 2019 are to expand in different geographies and in verticals other than LEA and Intelligence agencies. In the future, they intend to create an AI framework which can be used by individuals, and not just organisations, to analyse their own patterns and optimise their daily tasks.