A Startup Which Aims to Power Autonomous Vehicle Tech Like Never Before
iMerit has ten Centers of Excellence for AI training data in India, Bhutan, and the United States.
iMerit provides high quality data annotation across computer vision, natural language processing and content services that power machine learning and artificial intelligence applications for large enterprises across the autonomous transportation sector. Experienced iMerit teams extract intelligence from 2D and 3D images and videos to power the development and deployment of autonomous vehicle technology. During an interaction with Nitisha; Radha Basu, Founder & CEO, iMerit elaborates on the sectors they are catering to and also share her future plans to serve data intelligence and AI technology solutions to the top clients.
How did iMerit come into being and what are the different services that you provide?
iMerit was launched in 2012 with the intent of doing business while also making a positive social impact. Khosla Impact, the Michael and Susan Dell Foundation, and the Omidyar Network contributed US$3.5 million in Series A funding to iMerit in 2015. It received US$20 million in Series B funding in 2020 from British International Investment and its original investors. It now has ten Centers of Excellence for AI training data in India, Bhutan, and the United States. With over 5,500 employees, it has grown into a tech-enabled data annotation powerhouse.
At iMerit, we have a vibrant culture, with 54 percent of our staff being female. The average age of our employees is 24 years. They believe AI is a crucial part of our future and are eager to learn.
iMerit provides expert-level data labeling ensured by a rigorous learning and development process and industry-specific personnel. iMerit’s end-to-end AI data solution, which has tagged hundreds of millions of photos, provides support throughout the AI development cycle. In comparison to our competitors, we provide a full-time in-house annotation workforce, industry subject matter experts, unique solutions, ISO 27001 and SOC2 Type II compliance, in-house personnel training, and edge case expertise. We’re one of the first firms to offer completely integrated end-to-end AI data solutions for businesses using expert-in-the-loop machine learning workflows. It provides a comprehensive, end-to-end solution. Ground Control, Edge Case, and People Platform are all part of the iMerit Data Studio, which was created in response to clients’ requirement for rapid scale-up and deep knowledge in higher quality data.
Which all sectors do you cater to and which is the most upcoming one which is critical for iMerit?
We are at the forefront of the technologies working on autonomous mobility, medical AI, AI in precision agriculture, geospatial applications, commerce, social media, financial services and government applications.
iMerit is a global leader in providing technology services for autonomous mobility. We are proud to work with three of the world’s top five AV companies. We drive machine learning and AI solutions for leading organizations in the autonomous mobility sector by providing high-quality data for multi-sensor fusion, street mapping and localization, prediction analysis, driver assistance, and vehicle monitoring.
We have also worked with top healthcare firms on digital pathology by annotating cells and tissues on Whole Slide Image (WSI), and our digital radiology teams have annotated tens of thousands of plain films, CTs, and MRIs. Global leaders in robotics and endoscopy rely on iMerit’s expert computer vision data teams to drive innovation in instrument tracking, lesion detection, and phase identification.
The need for data annotation and AI-led solutions is a must in the planning, developing, and implementation of sustainable infrastructure for the autonomous sector. What is your take on it?
Many countries are recognising the rising need for and possibilities of autonomous technology and developing steps to support its development. For example, the United States recently passed a $1 trillion infrastructure package that includes a number of recommendations for updating infrastructure in order to allow wider adoption of autonomous vehicles and mobility. India is at a nascent stage but striving with great potential to explore AV technology. The number of startups and organisations developing AV technology around the globe are increasing.
As a global leader in data annotation for autonomous vehicles and having enhanced more than 150 million data points, we affirm that data is critical for AI models to work precisely. Since autonomous vehicle development is largely a visual interpretation task, all the training data is some form of video – from still images to full-motion video. When poor weather conditions, damaged signs, and insufficient contrast test the camera’s capabilities, other sensors like lidar and radar can help. Hence, collectively taking inputs from Radar, Lidar, camera, and ultrasonic sensors i.e multi-sensor fusion is vital to improve autonomous vehicle safety.
The push for AVs has led to massive innovations and driverless cars are already out on some roads, revolutionizing travel. To continue this rate of development, innovators must focus on high-quality data and properly deploy artificial intelligence (AI) models to foresee and assess edge cases, in addition to infrastructure, which is critical to the success of AVs.
What does the future look like for autonomous vehicles in India?
Although fully automated vehicles may be a distant dream for India, the market has witnessed some levels of automation. The post-pandemic focus has shifted to environmentally benign, self-manufactured, and long-term solutions. More and more consumers are expressing interest in electrified and hybrid vehicles.
While the demand for mobility is high, there are other innovations impacting the sector. Driven by the experience of the pandemic, the adoption of connectivity features has been the key. This is shaping not only personal mobility but also commercial vehicles, especially in the agriculture and infrastructure sectors to provide ease of business, safety and faster logistical processing.
There are still some challenges that make AV technology difficult to implement, especially in India, where roads are perilous, and traffic is unpredictable. Another source of concern for the government, which inhibits foreign investment, is the prospect of job losses due to automation.
The necessity of trial and error for India’s autonomous transportation business has been highlighted by recent start-up disruptions. Because data is the most important component, it is critical to ensure that all boxes are checked before proceeding.
How was your experience in the COVID-19 pandemic and how did you cope with it in terms of business and talent?
The pandemic was a particularly difficult time. Late in March 2021, all India and US facilities (90 percent of capacity) immediately shut down. The most significant change was switching our delivery method from 100% on-premises to 100% work from home (WFH). The difficulty was exacerbated by the fact that the majority of employees come from low-income households with inadequate infrastructure and live in neighbourhoods with limited connectivity. As a result, the teams worked non-stop to activate over 2,400 people who were under strict lockdown by distributing computers, installing broadband/dongles in over 1,500 homes, and collaborating with local ISPs to improve connectivity in low-income areas. Furthermore, the employees’ families and communities were made aware of the mission and specific challenges of working from home full-time. During the changeover, team leaders assisted them while managing their own difficulties.
In the last year, we grew 80% of our business in medical AI, applications in speech conversation, computer vision, object tracking, drone deliveries, geospatial as it applies to weather tracking for floods, and more.
In the year of COVID, we added about 1300 jobs. For skilling, you must now invest heavily in technology. The data solution from iMerit combines technology, talent, and techniques to deliver high-quality data and precision at the scale required. We meet the need for a consistent supply of data throughout the AI development process, from training to validation and deployment.
The Tech industry in India is tackling issues around employee retention and attrition rates. What are the different initiatives iMerit is undertaking for employee wellbeing and talent retention?
Hiring in iMerit is always a function of multiple key stakeholders being involved in the process or representations across functions being present in the panel, and a 100% agreement by all panelists in each hire is a mandate to ensure that any potential individual or functional level bias in the process is eliminated.
iMerit uses recruitment software Greenhouse (Greenhouse.io) and has integrated Greenhouse into its website (www.imerit.net/careers). Greenhouse is a SaaS solution that streamlines the entire recruitment process, both from an employer and candidate perspective. It is an applicant tracking system that allows us to manage large pipelines of candidates across every department and increases the speed of our overall process by enabling clear, consistent and immediate communication between hiring managers and key influencers.
Success of the people planning, and talent matching function is reflected in our retention metrics: iMerit has maintained over 90% retention over the past seven years which is unusual for a technology company.
iMerit PeoplePlatform is another continuous integration and deployment infrastructure for scaling operations. The platform is critical to optimizing task workflows and skill-matching expert workforces to projects at hand. The workflow management solution designed for iMerit facilitates the entire project lifecycle. iMerit PeoplePlatform enables project scalability while maintaining the quality of skilled resources.
At iMerit we invest in our employees in a variety of ways, including job creation, income stability, physical and mental health, and most importantly, skill development. Aside from the necessary technical knowledge, the Learning Academy at iMerit covers topics such as communication, comprehension, culture, and confidence to upskill employees and build the tech-propelled, advanced workforce of the future.
We provide a leadership programme that seeks to equip team leaders with the necessary abilities for managing people, projects, and successfully communicating in order to assure their personal development and the organization’s overall progress.
At iMerit, we are always finding opportunities for the overall wellbeing of our employees, catering to the needs of our diverse workforce. While in 2020, we focused on providing vaccination and health insurance primarily for Covid-19, in 2022, the focus is more on health awareness, and managing the hybrid work model. “My Covid Stories” was published as an initiative to boost the morale of our employees by telling them inspiring and important messages from iMerit’s Covid warriors. We also tied up with Truworth Wellness Platform to provide a host complete range of wellness for all the employees and their families. This platform is an app-based one that makes access easy as an anytime anywhere facility. This app has been of great help to employees as they can consult specialists and get their medicines delivered at their doorsteps while they are working at home. During the lockdown, “How to Work from Home” modules were created and delivered in ILT format to cater to the entire workforce. Additionally, virtual training sessions were introduced through the iMerit One learning platform and the creation of micro-modules was undertaken to cater to the continuous training needs of the staff.
What are your future plans for the company’s growth/expansion in India?
As the AI market moves closer to production and deployment, iMerit’s work and growth will reflect the changing priorities of teams working to leverage ML across industries. The idea is to really grow the company to be at the forefront in bringing this data intelligence and AI technology solutions to the top clients. We want to get AI into production, which is called ML Ops; we want to be the leader in ML Ops. We wish to be a Data Operation Center for all the data coming in and the analytics and insight from that.
What makes AI really difficult to go into production is the fact that after you have the models, you have a large number of edge cases; that is where human intelligence becomes so critical.
