Cult.fit provides a range of fitness-related services through a mix of online and offline channels. The company, founded in 2016 by Mukesh Bansal and Ankit Nagori, offers online personal training, group workouts, and live fitness classes across formats at Cult.fit centres, partner gyms and fitness centres across the country.
“While technology was always at the core of what we do, the pandemic fast-tracked everything – we rolled out our online fitness and wellness offerings which are still doing well. Before the pandemic, all our class schedules, workout plans, attendance management, etc. were highly automated. They still continue to be heavily reliant on technology for seamless flow of information and for precise scheduling that helps us avoid overcrowding and manage the flow of customers,” said Ankit Gupta, head of product and engineering, Cult.fit.
In an exclusive interview with Analytics India Magazine, Ankit shed light on the tech stack that empowers Cult.fit.
AIM: How does Cult.fit leverage technology to offer fitness solutions?
Ankit Gupta: Technology is the backbone of our live fitness offerings. From live sessions to pre-recorded classes, live class rankings, energy meters, and performance reports – our backend tech and powerful AI mechanism are largely responsible for managing the platform and handling daily traffic. We have been continuously working on improving our platform to ensure that the workout experience for users is as seamless as possible especially as more people have adapted to online fitness in a large way owing to the pandemic.
Beyond posture correction and rep counting, we are using AI to create customised workout plans for our users. Our tech DNA has been a major catalyst in our evolution. As a digital-first company, we constantly endeavour to push the boundaries of what tech can do to create a highly differentiated user experience in fitness and wellness, and our products are a clear example of that.
Our goal is not only to improve our operations at the backend but to also provide a one-of-its-kind fitness experience to our users. Broadly, we use AI for the following:
To create tech-led products and experiences:
1. Live fitness classes with energy meter: When a user works out using the Cult.fit app, he/she gets to measure performance via a live meter which tells how much energy has been spent during a particular session. It also tells the user where they stand in comparison to other attendees and helps generate an overall comprehensive report that the user can check to benchmark progress. The data that generate these insights is processed on the user’s device. The video feed never comes to our servers.
2. Smart workout equipment like cultbike and the upcoming cult rower: Immersive experiences in the virtual world are helping us give a boost to the smart fitness equipment we offer within the cult ecosystem. We leverage AI to create these life-like scenarios and virtual pathways that are engaging for users.
To run the business better and manage organisational operations: A lot of our backend business operations use AI. This has helped us streamline our operations and give better tools for our trainers, doctors, etc. to work with. These include:
· System-led class schedules based on micro-market level demand patterns
· System-led trainer allocations based on individual trainer constraints and business constraints
· Computer vision-based centre and class audits
· Prescription recommendation to doctors
To improve customer experience:
· Smart Workout Plan: We have launched version 1 of an offering called the Smart Workout Plan which is generated by an AI trainer. Through inputs received on a user’s preferences, basic information such as height, weight, fitness goals etc. our AI is able to curate a customised workout plan for the user. Today it also understands whether the user is someone who can work out daily or who can work out only on weekdays or three times a week. It also understands what their past performance has been. The AI, moreover, tweaks the plan for users based on continuous feedback. As we test and try the offering more, our AI continues to become more intelligent.
· App-led class bookings: This allows us to manage the flow of members in our centres and avoid overcrowding.
· Real-time feedback: AI help us elevate our engagement with users by giving real-time feedback – this is a major catalyst leading to improved performance by users in their fitness classes.
AIM: How is AI transforming the fitness industry?
Ankit Gupta: People have really busy lifestyles and prefer solutions that take away the brunt of conducting menial tasks on a day-to-day basis. Thus, AI-powered solutions have become a lot more popular because they are able to anticipate and solve quickly for common problems. The fitness industry has benefitted immensely from this – more so in the last 2-3 years.
Globally, reports suggest an increase of 46% in downloads of fitness apps post-pandemic. Clearly, people are looking at fitness a lot more seriously now than ever before – and they are turning to digital modes for this. Hence, this is an opportunity for AI-powered services to add a higher degree of customisation based on data at their disposal.
Fitness and health are unique to each person. A workout that works for me may not work for someone else. AI can solve this and build a level of differentiation within platforms that makes working out online just as, if not more, engaging that the traditional ways.
AIM: Tell us about your tech stack.
Ankit Gupta: Our tech stack includes the following –
· Cloud infrastructure – Mostly on AWS
· Compute Infrastructure – We use Kubernetes clusters on top of Spotinst
· We use stateless Java and Nodejs-based microservices which run on top of MongoDB, Redis and MySQL as a database.
· For analytics, we use Metabase based on Redshift data warehouse
AIM: How does Cult.fit ensure customer data privacy?
Ankit Gupta: Since we are a health and fitness company, privacy is a huge concern for us. We often handle data that is sensitive and, as a result, it is our duty to ensure that best practices are followed to maintain that aspect of data privacy. We adhere to all standards to enable this, and we are also undergoing a more rigorous certification so that we can also put that standard against our products.
Some of the ways we are ensuring data privacy are:
· We anonymise all user data before engineers and data scientists can look at them.
· Access is typically temporary and, on a need-to-know basis. We have a stringent approval process in place and only those who need to know certain information, such as doctors for consultations, get access to that information – no one else does.
· Most internal tools don’t show the personally identifiable information (PII) for the customer. Any request to such data in exceptional cases is logged and available for audit.
We have ensured that the data we collect is only on a need-to-know-basis and even then, we give a clear choice to our users. For instance, at Cult.fit, our features such as the energy meters are completely opt-in. People can choose to keep their cameras off during a session. However, even for users that do opt-in for camera access, we have remained true to the practice of never actually collecting any of the video feed. We knew while making this product that privacy is a major concern, and we respect that completely. All algorithms are operated on the user’s phone, none of which comes back to us.
AIM: What’s your long-term vision for the company from a tech perspective?
Ankit Gupta: Our long-term vision of the company from a tech standpoint is to marry sensory knowledge from the human body with behavioural techniques to create a truly personal health solution for everyone. There are three fundamental problems in health and fitness:
· People don’t understand the body the way they should. There are hundreds of metabolic processes and thousands of biochemical responses that we can’t track in a convenient and real-time manner. Tech and wearable monitoring devices can help bridge this gap in the long term.
· Everyone struggles with discipline and motivation when it comes to leading a healthy life. There has been massive innovation in the last year to help people with behavioural hacks and techniques to pursue their long-term goal(s). Having the right set of prompts, tech-led guidance and atomic habit techniques will be key in achieving our vision of making health easy for everyone.
· Most traditional forms of consuming fitness are boring or inconvenient. Tech can play a huge part in the creation of new fitness formats with built-in gamification and social features that are really fun to consume and at the same time do not require people to painfully cross convenience barriers.
