BLOG Inside Playment

2020 In Review: Looking Back at Playment’s Journey and Looking Ahead into 2021...

Siddharth Mall
January 13, 2021

We can all collectively agree that 2020 was an unexpected year for the world at large. And it is almost too easy to dwell on the negatives during such a large-scale crisis. But in retrospect, we cannot deny that this tumultuous year has helped us grow immensely. We took charge of the situation, consistently exceeded customer expectations, and used opportunities to stay true to our mission of expediting the AI age. We also found new ways to connect and collaborate with teammates virtually as we adjusted to the new remote working setups. 

Long story short: We managed to find a way to thrive in the new normal. Here's a quick recap of our year and the lessons we learned along the way. 

To a great start and successful partnerships!

In January, we were at the CES 2020, Las Vegas. We had an incredible opportunity to partner with Microsoft and showcase our data labeling capabilities along with the Azure ML team. Throughout the year, we have had many pilot successes, and we managed to bag long-term partnerships with global ML teams from companies like Samsung, Intel, Siemens, Sony, Dell, BMW, Bosch, LG, ZF, and Solera.

Playment Founders at CES 2020, Las Vegas.

Growing applications of computer vision and some interesting use cases we cracked this year… 

In 2020, we saw an increased rate of AI adoption by different governments worldwide and enterprises keen on automation. While we started out with a deep focus on autonomous driving use cases, this year we expanded with other industries like agriculture, real estate, mining, and defense among others. We created datasets for some interesting CV applications like aerial imagery for tower site prediction, weed detection in agritech, industrial components counting solutions, stock-out detection and gesture recognition in retail, real estate mapping, auto text categorization, and more.

When times got tough, we got smarter!

The pandemic took over in March, and India announced a 2-month lockdown. Soon, all our operations were taken remote. We devised quick and efficient contingency plans to ensure frictionless services for our customers. Thriving in the new normal came with its own challenges, yet, we were well-prepared to go remote because of our web-based labeling platform and pre-established collaboration and communication protocols. From product to solutions, all hands were on deck. We executed more than 776475 hours of labeling and shipped close to 64M high-quality annotations remotely for our customers worldwide.

We took customer success to the next level, quite literally.

Customer Success at Playment

Our agility and customer-focused processes helped us deliver high-quality training data on time for all our customers, even during the global pandemic. We extended contracts and continued serving customers who had to halt data collection due to the pandemic. We made our contracts more flexible, improved customer communication cycles, and quickly adapted to their sporadic labeling needs. After all, customer success is why we come to work every day!

We launched GT Studio Beta globally!

Playment’s GT Studio: A web-based data labeling platform for ML teams.

This year, we spent more than a few late nights developing a product that makes data labeling more accessible to ML teams. As more and more companies were going remote, we saw the need for a high-performing web-based labeling platform. And that is exactly what our product team delivered. A low-code platform that anyone anywhere can use to label datasets and manage 100s of labeling pipelines seamlessly. We also gave early access to a few of our customers and teams interested in exploring a data labeling platform that’s perfect for remote setups. We plan to roll it out publicly in 21Q1. More announcements to follow here.


| Learn more about GT Studio and here.

We also made data labeling easier — via ML automation features.

New problems call for innovative solutions. As labeling needs became more sporadic, timelines also started becoming shorter. And we are all aware that human labeling is a time-consuming process. That’s why we introduced ML automation features that reduce the time taken for labeling. We shipped many cool features like one-click cuboids, default dimensions, ML proposals, etc. We also perfected interpolations and other new features for video and sensor fusion annotations. Our human annotators label faster by correcting near-perfect annotations created by our tools. We divert their time to improve the quality of our deliverables with various QC tools and processes. 


| Read all about our automation features, workflow builder, and sensor fusion tools.

We got a makeover! Check out our new website… 

We have built a new website with first-of-its-kind information to give our community a peek into what really goes behind building data labeling infrastructures that produce high-quality training and ground truth data. 

We made it to the Forbes 30 Under 30 Asia list!

Akshay Lal, Ajinkya Malasane, and I were featured in the Forbes 30 Under 30, Asia list in 2020 under the Enterprise Technology category. Since the inception of Playment, our mission has been to expedite the AI age by democratizing access to high-quality, diverse datasets. Receiving this recognition is a testament to our belief in the promise of AI in the upcoming years. 

Playment also turned 5 in 2020...

Last September, Playment completed 5 years. We are more than grateful to our customers who believed in us and continue to partner with us. Most importantly, we would not have made it this far without a few key employees who have completed four years or more with us and helped us build this high-yielding data labeling platform we have today.

To celebrate the occasion, we hosted a virtual fireside chat with the veteran Playment employees and the Founder families to celebrate this occasion. The event served as a much needed motivational boost during the pandemic for all of us at Playment.

Thank you, @HR, for sprinkling some fun into a year that needed it the most.

From fun sing-off sessions to virtual heists and group competitions, our time working from home was not-so-dull. All credits go to our excellent HR Manager, Sahiba Chandok, who has created an inclusive and fun work culture at Playment. 

Lastly, cheers to phenomenal teamwork!

None of this would have been possible without the people working behind the scenes. From taking pay cuts during the pandemic to giving more than 100% effort for all our projects, we could not have asked for a better team. A big shout out to the Playment team! We are smarter and stronger together. 

We are excited about the promising future of data labeling in 2021 and beyond…

The next decade is going to see many crucial AI innovations. The focused ML efforts of 2020 will fuel the emergence of self-driving cars as early as 2021. Nuro, Zoox, and Lyft have already made announcements, and we are sure there is more to come. Adopting computer vision for camera surveillance of COVID-19 patients and increasing investments in AI-led research for vaccines in the healthcare sector has opened up many new opportunities for data labeling. We are excited to see the increased pace of CV adoption in several sectors. 

In the autonomous driving industry, lidar companies like Luminar, Innoviz, Velodyne, and Ouster have IPO-ed. This has strengthened our belief in lidar-led innovations. Lidar sensors available in the new iPhones have also bolstered the adoption rates. Playment will continue to build cutting-edge tools for lidar sensors in the upcoming year.

Since the future of work is remote, there is an increasing demand for web-based annotation tools compared to offline tools. Managing 1000s of annotators every day requires an analytics-led approach in the remote set up. Low-code infrastructure is the need of the hour. 

Playment has been proactively working on creating infrastructures that offer maximum flexibility and scalability. Our platform also provides robust productivity analytics that can be very useful in remote setups. The global launch of GT Studio will further democratise the access to high-performing labeling tools for ML teams worldwide. 

We’re gearing up for 2021 and you have a lot to look forward to!


This is just the beginning, there’s so much more to come.

Signing off!

Siddharth Mall, 

Founder and CEO, Playment