RecoSense Labs uses AI to track user behaviour, personalise content for them


Bengaluru-based startup aims to achieve higher user engagement, watchtime, loyalty and participation of users on online content platforms

While audience mean everything in the digital world, converting online viewers to loyal customers and getting them to buy products is a big challenge.

With a focus on intelligent data and engaging users with personalised content, RecoSense Labs was launched in 2014. The idea was born when companies like Netflix, Amazon and Google were winning customers globally by offering personalised technology.

RecoSense Labs, a SaaS-based platform, works with online content platforms and uses Artificial Intelligence (AI) to track user behaviour. It then interprets content suitable for a viewer and prolongs user engagement with personalisation.

The Bengaluru-based startup aims to achieve higher user engagement, watchtime, loyalty and participation of users on content platforms.

Amith Srinivas, founder of RecosSense Labs

The product and technology

Online content platforms invest a lot of effort and capital on acquiring users and building content for them. But sometimes they witness a drop in the number of users.

“When there is a steep drop out of users, the RoI (return on investment) on marketing goes down tremendously,” says Amith Srinivas, founder, RecoSense. He says the problem is to engage users on multiple channels. Developing a technology for personalisation is also effort-intensive and time driven.

“There is a strong need to simplify the automated personalisation by using data intelligence to achieve better user retention and higher number of transactions,” says Amith.

To achieve this, RecoSense’s product integrates with APIs (application programme interface) of the content management system of the customer, and the product also works as lightweight SDKs (software developer’s kits) for the front end – web, apps and devices.

After it collects data on the user behaviour, the AI-based product personalises content and product offerings by understanding content semantics. The company creates a knowledge graph, which uses Natural language processing (NLP) and Machine Learning (ML) to understand any key entities present in the content (such as people names, attributes, keywords) and categorise the content on multiple levels (in terms of topics, high level categories, etc). The system then builds a user profile with behavioural patterns to do three levels of mapping – user to product, user to user and then product to product.

The founder

RecoSense Labs is the brainchild of Amith Srinivas, who has earlier worked in data engineering solutions with companies like Mindtree and Harman. He worked for close to 12 years in the corporate world before stumbling upon the idea of using AI to track and target customers.

“We realised there were close to half a million digital properties who required the same technology to market and affordable cost of leveraging it without having to make an investment on building such a technology,” says Amith.

Market and Revenues

RecoSense offers its products to customers on a SaaS-based subscription model. It follows a monthly subscription fees based on the volume of users and premium fees on advanced features.

The company has raised an angel round of funding so far. RecoSense declined to disclose its clients, as it says it has signed a non-disclosure agreement with them.

According to Gartner, the global business value from Artificial Intelligence is projected to reach $1.2 trillion in 2018, an increase of 70 percent from 2017. AI-derived business value is forecast to reach $3.9 trillion in 2022.

Challenges and differentiating factor

RecoSense says global expansion was a huge challenge for them. “We have seen strong traction outside India, but we need to consistently build our visibility through investment on the marketing and sales team,” says Amith.

For now, the company is building a network with channel partners like technology platform integrators, solution developers and system integrators.

ReconSense says it has built its own tech stack on Python – a dynamic computer language - and open source components. The AI framework is also built on natural language capabilities (human languages interacting with technologies like Machine Learning) to understand regional content.

The startup sees competition from companies such as Reflektion, UnBxd, Rich Relevance, Qubit, Euler Systems and Webscale Networks, who are based in the US. In India, this business is a large opportunity as many brands have their own ecommerce and content websites.

According to the Ministry of Commerce, India has one million SMEs and each of them can use AI to win customers visiting their website. The AI understands the behaviour of visitors and throws relevant content to make them become potential clients.

RecoSense says it stands apart from its competitors with its Intellectual property (IP). “This IP becomes a key USP because it enables us to provide in-depth personalisation based on a comprehensive understanding of each content and hence every user,” says Amith.

For example, when a user views a pair of ‘Crocs’ online, the RecoSense IP can gather information that the user is currently interested in footwear and might have an affinity for comfort specifically. So, the AI concludes that the consumer might be specifically interested in a particular colour footwear, and is looking for a price point of less than $200. This interaction becomes a data set and the consumers are given the content based on this.

The team sources data from offline channels such as customer care and point of sales (PoS) data.

Future Plans

RecoSense plans to scale its business in the global markets and invest on sales and marketing in the next 18 months.

It is also betting on platforms that can enable personalisation of any kind of content – videos, text articles, ecommerce products and travel packages. “This gives us the opportunity to work with multiple domains,” says Amith.

“AI is the future and it replaces manual data gathering and understanding of the context. It allows scale because it throws in several attributes about the user,” says Amar Chokhawala, Founder of Reflektion.

For now, RecoSense is focused on winning more clients and scaling up its business.


Updates from around the world