iMerit combines the technologies, processes, and people that, together, deliver the data enrichment and annotations companies rely on. Today more than 5,500 iMeriters in the United States, Europe, Bhutan, and India work with many of the most innovative companies in the artificial intelligence and machine learning ecosystems, to unlock the value hidden deep within large volumes of unstructured datasets. Radha Basu is the Founder and CEO of iMerit and Analytics Insight has engaged in an exclusive interview with her.
Highest Quality End-to-End Services
iMerit is the leader in providing high-quality datasets for machine learning algorithms and artificial intelligence applications. Our services support all forms of data labeling and enrichment, including computer vision, natural language processing, transcription, and other content services. iMerit takes unstructured data or mediums and annotates them to create highly accurate structured data for advancements in machine learning and artificial intelligence.
Our tool-agnostic and inclusive approach ensures the correct tool is chosen for the task. Depending on the client’s needs and goals, iMerit can use our own data annotation tools, the client’s tools, or any third-party tools. Choosing the right tool for annotation and labeling is critical to the success of the project, both in terms of process, efficiency, cost, and data quality.
To ensure the highest quality data is achieved, iMerit provides end-to-end service to guide data scientists through all phases of their data annotation project and architect a solution that meets their needs. iMerit’s solution architects draw upon their deep expertise in all forms of data annotation to help define the requirements, training, workflow, feedback loops, and data validation required. Once the annotation and delivery process begins, they continue to use a dynamic improvement model by incorporating feedback through constant monitoring and edge case analysis to gain insights and further ensure data quality and accuracy.
Types of Data Labeling Supported:
• Computer Vision Services
Video and Image annotation
Keypoint and Landmark Annotation
Bounding Box Annotation
Lines and Splits
Lidar Data Annotation
• Natural Language Processing and Language Services
Audio Validation and Transcription
Sentiment and Intent Analysis
Named Entity Recognition and Linking
Topic or Object Classification
Grammatical and Discourse Analysis
• Content Services
Data Enrichment, Verification, and Extraction
Data or Content Quality Control
To provide high-quality, tech-enabled data services that leverage human intelligence to power algorithms in Machine Learning, while creating positive social change through employment in the digital economy.
To be the world’s most innovative data enrichment and annotation company, collaborating with clients and partners to deploy AI by unlocking the troves of value hidden deep within large volumes of unstructured data.
In 2015, the company raised US$3,500,000 in a Series A round of funding from Khosla Impact, Michael & Susan Dell Foundation, and Omidyar Network.
- In 2020, iMerit raised a Series B funding round that totaled US$20M, led by the CDC Group.
- Today the company has a global presence with ten delivery centers in India, Bhutan, and the United States.
iMerit CEO and Founder Radha Ramaswami Basu have been a pioneer in the Indian outsourcing ecosystem for the past several decades. In 2007, she co-founded Anudip Foundation, a non-profit focused on creating digital livelihoods at scale for youth from India’s disadvantaged communities. In 2012, she founded the for-profit iMerit as a place for Anudip graduates to find employment.
A Journey that Started from a Single Room
Under Basu’s leadership, iMerit became a tech-enabled data annotation powerhouse with over 5500 employees. Its mission of human-empowered computing finds reflection all through the organization, be it in the skilling model or its operational structures. iMerit’s model centers around providing employment in underserved regions, and harnessing latent talent in rural communities, through continuous and inclusive skilling.
With little in the way of a set roadmap, Basu and her team of innovators tested different approaches and created a for-profit model that is successful by any measure. It has brought financial stability and social impact to communities and families alike. iMerit takes a double-bottom line approach to its business, where social and financial impact metrics are enshrined in the company’s articles of association alongside normal business metrics and are tracked and reported quarterly. The company works with the world’s most innovative companies tackling projects in artificial intelligence and machine learning. Long-term relationships are built with pioneering organizations that place iMerit’s stable, skilled, and agile workforce at their core.
iMerit, which started in a single room, has now expanded to ten centers, with over 50% women. iMerit has continued to increase the size of its workforce during CoVid.
Basu, through iMerit, aims to fill a void in the billion-dollar data annotation and labeling industry, for which the most important skills today are in the computer vision and natural language processing sectors. A large number of people are required to work with a large amount of data and for that, she has successfully created a workplace that provides steady employment in local communities in underserved areas. A critical component of iMerit’s success is being able to bridge the gap between skilling and employment. By doing so, Basu and her team have opened up opportunities across different geographies and set iMerit employees on a path of professional growth and economic stability.
A Dynamic Leader
Radha Basu is the founder and CEO of iMerit, a global AI data solutions company delivering high-quality data that powers machine learning and artificial intelligence applications for Fortune 500 companies. She is a leading tech entrepreneur and a pioneer in the Indian software business. Under her leadership, iMerit has employed hundreds of skilled and marginalized women and youth in digital and data services worldwide. Additionally, iMerit employees contribute to growing industries ranging from virtual/augmented reality to the sharing economy to e-commerce and financial services.
Previously, Radha was the SupportSoft Chairwoman and CEO. She led the company through public offerings and built it into a worldwide market leader in support automation software. Prior to SupportSoft, she spent 20 years at Hewlett Packard, where she assisted HP’s electronic software division business and founded the company’s operations in India, and set up the first software center of any multinational in Bangalore.
Radha founded Santa Clara University’s Frugal Innovation Hub and co-founded the Anudip Foundation – iMerit’s sister foundation. She also serves on the boards of NetHope, Jhumki Basu Foundation, and the Miller Center for Social Entrepreneurship.
Radha Basu has a Masters in Computer Science and Biomedical Engineering from the University of Southern California and has done an Executive Management Program in Business Administration and Management, General from Stanford University Graduate School of Business.
She has received numerous awards including:
• India’s Best Leaders in Times of Crisis by GPTW India
• Business Leader of the Year by World Leadership Congress & Awards
• UN Women-ITU Gender-Equality Mainstreaming Technology Award
• Top 25 Women of the Web Award
• Silicon Valley Business Journal Women of Influence Award
• CEO of the Year
• Global Thinkers Forum Award for Excellence in Youth Development
Radha Basu’s Contribution
Radha Basu, through iMerit, aims to fill a void in the billion-dollar data annotation and labeling industry, for which the most important skills today are in the computer vision and natural language processing sectors. A large number of people are required to work with a large amount of data and for that, she has successfully created a workplace that provides steady employment and skills development in local communities in underserved areas. A critical component of iMerit’s success is being able to bridge the gap between skilling and employment. By doing so, Basu and her team have opened up opportunities across different geographies and set iMerit employees on a path of professional growth and economic stability.
2020 was particularly challenging with the Covid-19 pandemic which resulted in near-simultaneous, economic “shutdowns” in Europe, the United States, and India. The year also brought with it Cyclone Amphan in May, a catastrophic tropical storm that wreaked havoc in Kolkata, India, the city in which iMerit is based, and caused widespread damage in eastern India, the area in which the majority of iMerit’s delivery centers are based. In the U.S. the company was also forced to deal with Hurricane Laura — and a relentless hurricane season in the Gulf of Mexico — which greatly affected operations and staff in New Orleans.
During a period of extreme challenges, Basu led a transformation within iMerit that found it changing from a company that conducted 100 percent of its operations “on-premise” to 100 percent of its operations remotely. At a time when many companies were laying off and furloughing employees, iMerit was procuring laptop computers, working with broadband providers to lay cable in villages in India, upskilling their workforce, and hiring employees in India, Europe, and the United States.
Examples of Basu’s Exemplary Leadership Include
• Launching unique wellness programs during Covid to ensure employee health.
• Ensuring a strong connection between executive leadership and the global workforce through more than 100 virtual fireside chats with employees.
• Adding over 800 jobs in six months from July – Dec 2020
• Leading a company metamorphosis, from vibrant center-based culture to 2500+ people working from home in rural and remote areas.
After all the challenges of 2020, through Radha Basu’s leadership, iMerit ended the year stronger than ever, reaching new highs in client satisfaction, revenue, employee growth, and workforce skills development.
Innovative Solutions to Deliver Relevant Business Outcomes
iMerit transforms unstructured data into useful enterprise-grade information. Through annotation and labeling, we create highly accurate structured data for advancements in machine learning and artificial intelligence.
Data annotation enables transformative technologies to come to life, such as advancing cancer cell research through CT scan analysis, optimizing crop yields based on drone footage, and training driverless cars by annotating videos to understand their environment. By leveraging iMerit’s data annotation services, customers can focus on improving the applications that drive their company forward.
• Image Annotation
• Video Annotation
• Text Annotation
• Audio Transcription
• Data Annotation
• Sentiment Analysis
• Lidar Annotation
• Content Moderation
• Product Categorization
• Image Segmentation
The Company’s Biggest USP
iMerit provides end-to-end AI data solutions to guide data scientists through all phases of their data annotation project and architect a solution that meets their needs. The iMerit team of subject matter applies its knowledge to every phase of clients’ process, providing guidance through the process – which involves customized solutions based on client needs.
Experts in the loop work with the client through the entire lifecycle:
• Set-up phase: defining and refining requirements, and designing the workflows to achieve goals
• Execution phase: experts who annotate the mediums, covering various types of annotations and solving the problem presented to us by the client
• Refinement: Adaptive model built on this methodology that involves feedback loops, edge-case identification, and refinement to achieve the highest quality of data
• Analytics: Insights on process flow, efficiencies, quality controls
• Overcoming the Challenges with High SpiritsCOVID-19 Crisis – The company was put to an extreme test when all our India and US centers (90% of our capacity) were abruptly shut down in late March. The biggest shift involved changing our delivery process from 100% on-premise to 100% Work from Home (WFH). The challenge was even more because most of our employees come from humble homes with inadequate infrastructure, in communities with limited connectivity. Rather than accept defeat, iMerit’s team worked tirelessly to activate over 2400 people under stringent lockdown, by distributing machines, installing broadband/dongles in over 1500 houses, and working with local ISPs to upgrade local connectivity in very poor neighborhoods. Further, the families and communities of the employees were sensitized to the mission and unique demands of a family member working from home full-time. iMerit’s team leads helped their team’s transition while also managing their own crises during the pandemic.
• Retaining the quality and content of the skilling structure and providing multi-level training as necessary to each and every employee – Most employees are first-generation learners and do not have advanced academic degrees, retention of quality has been a challenge. Over 80% of our employees hail from impact backgrounds and most are new to the technology field. Training in linguistic, cultural, and behavioral skills was required for client communications. Gauging the grooming necessary and the adaptive capability of each employee and thereafter training them was a challenge for the company. A robust L&D structure and the use of technology and a custom Learning Management System have helped the company maintain the quality of skilling, and prepare our teams to work on cutting-edge AI data projects.
Disruptive Technologies to Leverage Business Development
The trends that we’re seeing in big data analytics and AI vary depending on the maturity of the product or service being developed. Teams in the research and development stage are looking for large structured ground truth datasets to validate their hypothesis and get their project off the ground. The sensitivity to speed is lower, as the project is in its early stages. Teams are seeking guidance to develop their frameworks and build a robust pipeline for model development with repeatable and explainable processes.
As AI teams move closer to production, they are more focused on finding and solving edge cases in the datasets. The speed of iterations becomes more critical and real-time time insights into the process are required. With a massive volume of data, workflow management and appropriate tooling become key.
A Road to the Future of Investments and Development
As the AI market moves closer to production and deployment, iMerit’s work and growth too will reflect the changing priorities of teams working to leverage machine learning across industries.
Over the next iMerit will make significant investments in technology to better enable AI data solutions across the lifecycle of products and services. The focus will be to develop and provide a complete package to clients, from a workforce and technology perspective. This will involve an investment in the company’s infrastructure and ability to scale alongside our clients. There will be a continued emphasis on solving complex challenges in ML DataOps. iMerit will invest in building a deeper understanding of tools that can enable creation as data moves from the unstructured stage to pre-ML, and all the way to production, where data management is required.
Thoughts on the future of IoT/ artificial intelligence/ML/cloud computing in India
AI in India is projected to grow at a CAGR of 20.2% and is expected to reach US$7.8 billion by 2025, according to a recent IDC report. There is increased activity and investment in automation and intelligent solutions across sectors, including healthcare, agriculture, retail, and social media. Many customer service applications are seeing traction, a natural use case to tackle the massive volume of queries in any sector.
Focused research across India’s academic institutions will be required to build a strong ecosystem of experts to power the growth of AI. While leading engineering universities like the IITs and IISc have invested heavily in machine learning research, there is a need for wider adoption and AI-specific skill development across the academic landscape. Technical talent is currently seen in limited urban pockets but for widespread AI growth, knowledge and expertise need to be distributed across states.
Government investment in infrastructure for AI, along with interest and investment from the VC community, has already seen an upward trend. The creation of policies and frameworks that support data sharing and usage, especially in sectors like autonomous mobility, should be undertaken by experts and lawmakers with a robust understanding of AI and its applications.
As these sectors continue to mature, data preparation will be foundational, particularly in sectors where digital transformation is still underway. With a variety of data sources that need to be digitized and structured for use, data labeling and the creation of ground truth training data will be crucial. The explosion of IoT has resulted in vast volumes of sensor data that can be leveraged for AI applications. Challenging annotation workflows like multi-sensor fusion and those that require subjective analysis of scenarios will be required to map the world around and build ground truth for AI applications.
1) In-house, Full-Time Annotation Workforce
iMerit’s 5500+ highly skilled full-time employees across the US, India, and Bhutan with over 92% retention rate makes it easier to ramp up data labeling projects without compromising on quality. It allows long-term solutions with a knowledge ramp-up curve as employees achieve deeper expertise on customer needs.
Our full-time data workforce is recruited from underserved communities and skilled to perform a range of data annotation tasks that require different degrees of specialization and subject matter expertise.
2) Industry Subject Matter Experts
iMerit believes that past industry experience grants credibility. Our subject matter experts in the autonomous vehicle domain have deep prior experience within the industry and utilize the knowledge to guide the client and manage the projects efficiently. iMerit combines the best of predictive and automated annotation technology with world-class data annotation and subject matter experts to deliver quality output.
3) Custom Solution Architecture
iMerit’s protocols, tooling, and workflow can be customized based on customer needs. We recognize that the data training process is iterative and evolving and we are always as agile and flexible as one needs us to be. A solution consisting of disaggregated stages allows for flexible solutions for clients.
4) Quality Control: Closed Feedback Cycle
A closed feedback cycle is built-in with real-time monitoring and service delivery insights. The full-time expert model allows for traceback and repeatability down to a granular level. We also employ an evaluation model that continually assesses deliverables, key metrics, quality control processes, and business outcomes.
5) Security: SOC 2, GDPR compliant, ISO certified
iMerit is SOC 2 Type II and GDPR compliant, ISO 27001:2013 certified, and has been audited based on AICPA guidelines. We manage over 4,000 full-time employees across the US, India, and Bhutan under strict NDA. For high-security work, we operate dedicated and monitored facilities with strict security protocols.
6) Multi-Tiered Workflow for Scalability
A structured approach to data labeling is essential when scaling a data pipeline. iMerit has developed a phased approach to working with clients that scale for any project and holds a frequent basis of communication as the gold standard of client-vendor collaboration.
Before any data labeling begins, iMerit first deploys solutions architects who work with a client’s engineering teams to develop a comprehensive understanding of the team’s goals, challenges, and scope. After this evaluation, the solutions architects create a workflow in tandem with the client’s engineering teams. This multi-tiered approach forms a clear, structured workflow that will direct the future and all communication of the engagement.
7) Tool Inclusive: In-House, Client, Partner Tools
The team at iMerit is tool agnostic, which means we leverage a range of tools to make data labeling as painless, easy, and fast as possible for our clients. iMerit has in-house tools for annotation, workflow management, and quality control. iMerit also works with client tools and has a robust ecosystem of tooling partners, including AWS Sagemaker, Appen, Dataloop, Deepen, SuperAnnotate, and more, which allows us to cater to any project in the AV space.
8) Tool Guidance and Solutioning
iMerit’s solution architects put together the optimal tooling workflow for each project in consultation with technology, delivery, and learning & development. iMerit experts can provide guidance on which tool will work best for the project, based on the type of data to annotate, the size of the project, the workflows of the team, and the scope of the budget. Endpoint customization using tool APIs and SDKs add to the capabilities which are available off the shelf.
9) Custom Tooling
Choosing the right annotation tool is critical to creating high-quality data for machine learning applications. iMerit offers custom tooling, in addition to its in-house tools and partner tools. We can design and build custom tools to suit the specific requirements of the client. It allows iMerit to be flexible and to customize solutions expressly and exclusively for customer needs.
10) Expertise with Edge Scenarios
iMerit’s proprietary team of experts and annotation professionals are leaders in the field of edge scenarios (an unusual occurrence for the AV to understand) thanks to their extensive experience and training.
iMerit takes great care to document each and every edge case encountered in order to continually anticipate them in future cases and help their clients overcome them. When working on large-scale data annotation projects, iMerit employs an expert team of data labelers who’ve worked on other autonomous vehicle projects in the past. The expertise these data labeling teams bring to the table includes decades of experience working with and solving edge cases, particularly through anticipating them based on past experience.
11) In-House Skilling
Naturally, our clients don’t want to depend on us for everything. We are ready to support scale, when, and how the client needs it. This includes adding in extra hours to working on weekends to spinning up quick teams for ramp-ups.
• Trained for different initiatives: iMerit labelers go through work from home training, domain-specific certification programs, tools and applications training, and soft skills training.
• Involvement of Learning and Development: We have our in-house trainers who take the project training forward using the materials shared. In this way, training new people during ramp-ups is managed, with the limited time required from the client.
• Buffer resources: We provide more than the resources asked by the client and keep them at the same level as the others by rotating within the team. That way, the client does not lose any hours despite absenteeism.