A look inside Portl’s AI smart mirrors powering personal fitness journeys

We strive to be ISO 27701, GDPR compliant, and HIPAA compliant.

Launched in March 2021, Hyderabad-based Portl designs and manufactures proprietary smart mirrors that work as a personalised and on-demand fitness trainer and digital health and wellness companion. Their flagship product Portl Studio comes with a 43′ multi-touch screen with advanced sensors and heavy-duty compute power. You can sync the Portl studio with your smartphone to get started on the fitness journey. 

Analytics India Magazine got in touch with Vishal Chandapeta, co-founder and CTO of Portl, to understand how the company embeds ethics into its AI-powered fitness mirrors.

THE BELAMY

Sign up for your weekly dose of what's up in emerging technology.
Vishal Chandapeta, co-founder and CTO, Portl

AIM: How does Portl leverage AI?

Vishal Chandapeta: At Portl, we leverage AI/ML in several ways, from movement recognition, movement analysis, predictive analytics based on physical and medical information, and recommendation systems. Our AI/ML is then used to give our clients suggestions on how to better themselves, whether through bespoke physical programs, real-time automated movement modification and coaching, nutritional intake, mental well-being, or lifestyle feedback.

AIM: Tell us about Portl’s AI governance framework.

Vishal Chandapeta: All our AI/CV development is done in-house with no third parties used, from data collection to training to implementation. This gives us good control over ensuring fair and ethical practices are followed, especially in the data collection process. Being a people-first company, we hold ourselves accountable to the highest standards regarding data privacy.

One example of our data collection process was the way we went about collecting image data for our pose estimation. We ran campaigns in universities and gyms across the country where we attended and took images with the full consent of people, as well as giving a small presentation about why the images were needed and how they would be used in our company, allowing for complete transparency during the data collection phase. We also provided complete anonymity and obfuscation of information of the individuals who contributed. 

AIM: What explains the growing conversation around AI ethics, responsibility, and fairness? Why is it important?

Vishal Chandapeta: The tech industry has been under a lot of scrutiny. Today, we see tech adoption across huge demographics, from children to senior citizens. So it is important to ensure that what is developed is done in such a way that it does not take advantage of people who may not be as tech-savvy as others.

Coming to AI, to trust such life-altering decisions to a purely mathematical system, no matter how advanced, should not be allowed. Instead, I would suggest using such tools to make recommendations, and we must allow the human element to make decisions. We must also ensure that ML algorithms being used are vetted by a third party to ensure transparency in a company’s AI/ML practices.

AIM: How does Portl ensure adherence to its AI governance policies?

Vishal Chandapeta: We follow an internal development transparency system. All implementations done by every designer are reviewed and vetted by other people at a higher logic level, if not at a code level. Moreover, we constantly keep ourselves informed of the latest developments in AI governance policies and implement the same, both at a company and personal level.

AIM: How do you mitigate biases in your AI algorithms?

Vishal Chandapeta: I believe this is a deeper question than what it seems like from the outset. How to avoid bias remains one of the most coveted goals of AI. If we say humans themselves are prone to biases, then it is expected that an advanced enough AI will also be the same. So the aim is not to remove bias from the AI but to ensure that the AI is fair, ethical, and inclusive, even with any inherent biases.

We strongly believe in FATE (fairness, accountability, transparency, and ethics) in our AI/ML implementation. We also keep ourselves up to date with what is happening in the AI/ML ethics space worldwide. Due to the rapidly transforming nature of AI and its implementation and the repercussions of unethical practices, we constantly strive to stay ahead of our competitors in maintaining fair and ethical practices.

We also have a highly experienced academic set of individuals who help vet, regulate and ensure all fairness and ethical practices are followed.

AIM: How does Portl protect user data?

Vishal Chandapeta: We strive to be ISO 27701, GDPR compliant, and HIPAA compliant. Portl treats all information with the utmost respect; as a user myself, I constantly ask, “what would I be comfortable letting someone know? Would I be happy if this was seen by another person?” and, as such, ensure that every iota of information received by us is treated with the utmost of privacy, security, and obfuscation, to prevent anyone other than the original user from seeing the information.

AIM: Did you come across any biases or ethical concerns/issues lately within your organisation/industry/product? If yes, how did you address them?

Vishal Chandapeta: Early on, we did, but these were expected biases, which occurred due to the limited data we had. Our team had expected to see such things, and we fixed such biases with continuous and diversified data collection to expand our datasets. As we move forward with our company and the amount of information we get increases, we continuously strive to improve our data collection processes while simultaneously reducing unwanted biases, allowing complete transparency to our end users and peers alike.

More Great AIM Stories

Sri Krishna
Sri Krishna is a technology enthusiast with a professional background in journalism. He believes in writing on subjects that evoke a thought process towards a better world. When not writing, he indulges his passion for automobiles and poetry.

Our Upcoming Events

Masterclass, Virtual
How to achieve real-time AI inference on your CPU
7th Jul

Masterclass, Virtual
How to power applications for the data-driven economy
20th Jul

Conference, in-person (Bangalore)
Cypher 2022
21-23rd Sep

Conference, Virtual
Deep Learning DevCon 2022
29th Oct

3 Ways to Join our Community

Discord Server

Stay Connected with a larger ecosystem of data science and ML Professionals

Telegram Channel

Discover special offers, top stories, upcoming events, and more.

Subscribe to our newsletter

Get the latest updates from AIM
MOST POPULAR

Will Tesla Make (it) in India?

Tesla has struggled with optimising their production because Musk has been intent on manufacturing all the car’s parts independent of other suppliers since 2017.

Now Reliance wants to conquer the AI space

Many believe that Reliance is aggressively scouting for AI and NLP companies in the digital space in a bid to create an Indian equivalent of FAANG – Facebook, Apple, Amazon, Netflix, and Google.

[class^="wpforms-"]
[class^="wpforms-"]