iMerit is a company that creates software-delivered AI data solutions. With 10+ years in enterprise AI data, iMerit works on client data to prepare it for training their AI models through a combination of software, domain knowledge, solutions consulting, and human-in-the-loop. Pulse 2.0 interviewed iMerit CTO, Sudeep George, to learn more about the company.
Sudeep George’s Background
“My passion is for solving complex challenges in systems engineering. I had a leadership role in Embedded Systems at Sarnoff, where I spearheaded the development of product systems leveraging their cutting-edge computer vision IP. This experience provided a deep dive into the full technology stack, from complex hardware to sophisticated algorithms, and fueled my desire to build impactful solutions.”
He added, “I then spearheaded product and engineering initiatives for high-tech imaging systems at Serial Innovations India. I also co-founded Tonbo Imaging, a pioneer in computational imaging. At Tonbo, we built a range of electro-optical imaging products for the military, security and commercial sectors. Witnessing the rise of AI/ML and cloud computing, I recognized the potential to transform traditional computer vision.”
“I joined iMerit as VP of Engineering in 2021. If data is the new oil, iMerit is the oil refinery. The chance to build and deploy advanced computer vision and multi-sensor fusion solutions for a company with an established and sophisticated client base was incredibly appealing. I thrive on creating cutting-edge technology that makes a real-world impact. It is especially exciting when your technology is in the hands of expert customers and you can see it being used at scale.”
Roles and Responsibilities at iMerit
“At iMerit, I balance immediate business needs with future innovations in an extremely dynamic AI landscape. We serve large clients who are also responding to these changes. We are also simultaneously navigating a constant influx of cutting-edge projects. This requires deep technical expertise in high-performance systems design and novel ML development.”
“Recently, I was appointed CTO of iMerit. As CTO, I oversee product delivery within our tech stack and foster experimentation with scalable, robust techniques. This necessitates close collaboration with the business side to understand the customer ecosystem. I dynamically manage resources and priorities to address both immediate client needs and long-term innovation goals. I also have to understand, experiment and leverage new technologies to further our business goals. “
“Unlike early-stage tech or R&D, our work has an immediate and tangible impact, requiring rapid scaling, customization and deployment. We already have dozens of running customers with multiple workflows at scale. For example, our 3D multi-sensor fusion tool, processing tens of millions of 3D points and pixels across hundreds of frames, is rapidly attracting new customers in the automotive and mobility sector. Automation is another key focus area. It’s a big part of the future of our company and customers are asking for it from Day 1. We use a combination of smart tools and AI driven automation to achieve more and more efficiency which the industry demands.”
“Finally, I also play a key role in communicating our capabilities to customers and stakeholders. Especially in such complex and niche areas, clear and constant information is very valuable to them in shaping the go-to-market and direction of the company as well as new conversations with potential customers.”
Favorite Memory
What has been your favorite memory working for the company so far? George reflected:
“My favorite memories at iMerit stem from witnessing the direct impact of our technology on both our delivery team and our customers. It’s incredibly rewarding to see how the right tools empower our team to deliver higher quality work with greater efficiency. You feel motivated when a smart tool saves 30-40% effort, for example, as we achieved in a large precision agriculture workflow of crop and weed detection.”
“My favorite memory is of a specific project where we had a global workforce of subject matter experts working closely with technology teams to rapidly iterate and deliver results for a customer facing a critical deadline. The seamless synergy between our talented experts, rapid development cycles, and collaborative validation across iMerit’s teams enabled us to deliver exceptional value to the customer within a tight timeframe. This success story exemplifies iMerit’s core principle: the powerful synergy between tools, talent and technique.”
Core Products
What are the company’s core products and features? George explained:
“Our flagship offering is Ango Hub, which is AI data workflow automation software. It allows customers to load their data workloads and design a custom workflow that mixes and matches AI models with our humans in the loop to achieve high-quality annotation. As part of the whole AI lifecycle, it can also be used to perform audit and checking of operating models. Accordingly, it has strong features for data pipeline integration, workflow design, Bring-your-own-model for pre-annotation, custom interfaces for human-in-the-loop annotation/QC work, and dashboards to manage these operations at large scale. Ango handles multiple modalities such as image, video, audio, text and Lidar.
Within Ango, we have vertical modules for mission-critical sectors that we work with.
- Automotive AI – Our high-performance 3D multisensor fusion tool is for automotive use cases where multiple inputs – Lidar, Camera, IR – are being correlated in each frame to achieve the best possible perception of a scene. This can get really heavy and requires strong caching and rendering performance to handle millions of lidar points in a sequence of hundreds of frames. We recently launched an Automotive AI center of excellence in Coimbatore with specialists trained on these workflows.
- Healthcare AI – We’ve been doing medical imaging work for a while now. Our in-house specialists and global network of medical experts leverage our radiology suite to handle DICOM imagery, multiplanar scans with 3D views and custom tools. Our domain insights come from the senior physicians on our staff.
- GenAI – in the past two years we’ve also built workflows for GenAI prompt engineering, fine-tuning, chain-of-thought, RLHF, and red teaming. ”
Challenges Faced
What challenges have George and the team faced in building the company? George acknowledged:
“Data annotation is a complex field. While our software is broadly full-featured across multiple modalities like video, image, Lidar and text, we do not force a one-size-fits-all solution on clients. We also build custom data pipelines (integration with Databricks, etc.), custom workflows and features, and domain-specific capabilities, for example Radiology, Multi Sensor fusion and GenAI workflows for Supervised Fine Tuning.”
“This means we get to delve into the core challenges faced by our diverse customer base, each tackling their problems with unique approaches. “
“The rapid pace of change in the AI sector, particularly in the last two years, presents a significant challenge. It’s not just incremental change. There have been huge shifts with weekly announcements shaking up the sector. Both customers and iMerit are constantly adapting and making critical decisions about future directions, leading to frequent shifts in priorities. Maintaining a stable product and technology roadmap for an extended period is difficult in this environment.”
“Another challenge stems from the multitude of pathways and frameworks in AI and automation. Evaluating these options often requires hands-on experimentation, as theoretical knowledge alone may not be sufficient to determine the best approach. This necessitates a hands-on leadership style, which can be challenging for those accustomed to traditional management roles.”
“It is always a change in mindset for engineering teams, from being motivated only by one’s own technology stack, to considering all aspects of the customer end-use. I don’t think this can be left only to the product managers. In such a technical field, engineers themselves have to be customer focused.”
“Recently, developing our 3D multi-sensor tool for autonomous vehicles has been a demanding but rewarding experience. This complex tool requires a powerful point cloud and video management engine, pushing the limits of existing libraries. Even minor inefficiencies in the UI or backend can significantly impact user experience and project success. This demands a strong grasp of core software aspects like caching and full-stack development, while pushing the envelope when it comes to developing high performance WebGPU based applications, requiring our team to continuously elevate their skills.”
“The transformative power of automation and scalable technology is seen on an everyday basis.Finally, bootstrapping automation to help workflows is hard without access to the customer’s data to train our own models (this is a policy of the company). We have to weigh the noise created by a preliminary model trained on public data, versus the benefits.”
“Despite these challenges, the direct interaction with users and customers provides invaluable insights, allowing us to observe, learn, and iterate quickly. Witnessing the immediate impact of our improvements is incredibly gratifying.”
Evolution Of The Company’s Technology
How has the company’s technology evolved since launching? George noted:
“iMerit’s technology has undergone a significant evolution. Initially, our focus on quality, efficiency, and consistency led us to develop technology that supported our internal operations and skill-building (iMerit People Platform for resource mapping and iMerit One for skills development). “
“In the next phase, we partnered with third-party software companies like Dataloop, Datasaur, AWS Sagemaker, to offer full stack solutions for our clients. We focused on developing robust reporting, analytics, and insights (RAI) capabilities to enhance these workflows. Our flexible approach, prioritizing customer needs over pushing specific software, has been well-received, positioning iMerit as a trusted advisor.”
“The acceleration of AI development cycles has recently driven demand for end-to-end solutions. Recognizing this, we launched Ango Hub in 2023, our own comprehensive software with specialized modules for areas like multi-sensor fusion and radiology. While Ango Hub is a core offering, we maintain an agnostic approach, integrating with existing customer tools or specialized partners when necessary. Our commitment to providing the “best stack for the job” has fostered trust with both partners and customers.”
“Currently, 70% of new customer revenue is generated through Ango Hub, and this figure continues to grow. Ango Hub’s ability to integrate deeply with customer data pipelines and embed custom automations and quality checks enables us to deliver superior outcomes.”
“It’s clearly become a cornerstone of our technology strategy. ‘Software delivered services’ is the mantra. It allows us to commit to end-to-end outcomes and build in efficiencies while constantly offering new domain centered features to expert customers. All our institutional knowledge gets captured in this.”
Significant Milestones
What have been some of the company’s most significant milestones? George cited:
“iMerit has achieved several significant milestones throughout its journey. Celebrating our ten-year anniversary in 2024 was a particularly proud moment as an AI company.”
“Securing our first $10 million client marked a significant step in our growth trajectory.”
“Personally, I found the development of a cutting-edge synthetic data solution for the agritech space particularly rewarding. By simulating plant and weed growth under various conditions, we augmented customer data for an agricultural solution, demonstrating the power of Generative AI in image processing. This work even culminated in a published research paper.”
“The launch of Ango Hub last year marked a significant shift towards software-delivered services. Ango Hub was our first major customer-facing software, driving a new level of engagement. Currently, 60% of new customers are acquired through Ango Hub. We’ve further enhanced its capabilities with a 3D point cloud module for multi-sensor object segmentation, which has been instrumental in securing several promising new accounts.”
“This year, we also launched iMerit Scholars, a curated network of expert consultants who provide domain-specific fine-tuning and alignment for Generative AI. This addresses the need for specialized expertise and fosters respectful relationships with highly skilled professionals.”
“Most recently, we unveiled ANCOR, our first co-pilot tool for radiology, at RSNA. ANCOR assists with navigating complex guidelines and improving the accuracy of radiology annotations.”
Customer Success Stories
When I asked George about customer success stories, he highlighted:
“Here are a couple of examples that highlight the power of our software-driven services and our emphasis on the synergy of talent, tooling, and technique:
Autonomous Mobility: A customer in the autonomous mobility space using a complex sensor suite and generating massive amounts of data needed a scalable annotation solution. They also wanted deep integration with their data and ML training pipelines for optimized development.
— iMerit’s Solution: We rapidly deployed a custom solution integrating Ango Hub with the customer’s technology stack. Our multi-sensor tool was tailored to accommodate their specific sensor suite, and the web stack, including WebGL and GPU components, was iteratively optimized to handle their unique data load. iMerit One, our training platform, was customized to enhance team skills in scene analysis and classification, and our RAI capabilities were leveraged to improve project execution techniques, increasing throughput and accuracy.
— Result: This multifaceted approach resulted in successful project execution and a satisfied customer.
Generative AI Fine-tuning: A customer sought to develop a specialized dataset for fine-tuning a Large Language Model for a niche use case.
— iMerit’s Solution: Leveraging Ango Hub’s flexibility, we rapidly created a customized RLHF (Reinforcement Learning from Human Feedback) tool that met their specific requirements. This tooling enabled our subject matter experts to successfully generate a unique corpus of data, delivering the value the customer expected from iMerit.
— Result: The customer received a high-quality, tailored dataset that effectively addressed their fine-tuning needs.
These examples demonstrate how iMerit combines technology, talent, and tailored solutions to address diverse customer challenges and deliver exceptional value.”
Funding
When asking George about the company’s funding details, he revealed:
“iMerit has Series A and Series B funding – a total of about $25 million. Revenue is growing, with our key verticals being Autonomous Mobility, Healthcare AI and GenAI for high tech. Additionally, Agricultural AI is growing fast.”
Total Addressable Market
What total addressable market (TAM) size is the company pursuing? George assessed:
“The immediate space, which is data in AI, is about $2 billion to $4 billion by 2026. The overall software delivered services space can be many times that amount, depending on which areas one takes on.”
Differentiation From The Competition
What differentiates the company from its competition?
“iMerit stands out from the competition due to a combination of key differentiators:
- Unmatched Experience and Institutional Knowledge: With a decade of specialized experience across hundreds of use cases and customers, iMerit possesses deep institutional knowledge. This is further strengthened by a 90% employee retention rate, ensuring continuity and expertise that’s hard to replicate.
- Leadership in Multimodal AI: iMerit’s extensive work across video, audio, niche sensors, 3D, medical imaging, and text data uniquely positions the company to excel in the evolving landscape of multimodal Generative AI.
- Seamless Software-Delivered Services: iMerit offers a truly integrated approach, blurring the lines between software and services. This comprehensive solution encompasses custom data pipelines, domain expertise, and a commitment to delivering exceptional data quality and outcomes. In AI, this is an essential capability.
- Sustainable Growth: iMerit has maintained profitability or breakeven for almost 4 years. It allows us to serve long-term clients and not lose focus.
- Curated Teams and Expert Networks: Unlike crowdsourced platforms, iMerit relies on a carefully curated and highly engaged network of in-house specialists and external subject-matter consultants, ensuring high-quality expertise and a tailored approach.
- Genuine Customer Obsession: iMerit’s customer-centric approach is deeply ingrained in its culture. By consistently prioritizing customer needs and building solutions that address real-world challenges, iMerit fosters long-lasting partnerships and trust.
These differentiators collectively position iMerit as a leader in the data solutions space, offering a unique blend of experience, expertise, and commitment to customer success.”
Future Company Goals
What are some of the company’s future company goals? George concluded:
“We aim to be the leader wherever human-in-the-loop is used to make AI usable. The role of human-in-the-loop remains crucial at various touchpoints, but it needs to be optimized for efficiency and accuracy. Designing an ecosystem that enables this at scale, while ensuring security and measurability, presents a significant challenge that we’re eager to address.”
“Ultimately, our primary goal is to remain laser-focused on the evolving needs of our customers. By consistently delivering exceptional value and anticipating their requirements, we believe that everything else will fall into place.”