Keymakr provides image and video data annotation, data creation, data collection, and data validation services for AI/ML computer vision projects. To create precise datasets, Keymakr developed Keylabs.
Michael Abramov’s Background
Could you tell me more about your background? Abramov said:
“I’m a software engineer with experience in data collection and a background as an R&D manager. Passionating about exploration and innovation, I always aim to make a meaningful impact in my work. Working with AI and annotation provides me with opportunities to engage directly with diverse industries — from automotive to agriculture — and play a part in driving their advancements and breakthroughs.”
Formation Of The Company
How did the idea for the company come together? Abramov shared:
“Keymakr and later Keylabs.ai were co-founded by Arie Zilberman, Michael Seldin, and me. For all of us, it was the right moment to respond to the growing demand for accurate training data for computer vision and machine learning. Almost ten years ago, AI and computer vision weren’t as ubiquitous as they are today, but we were confident in their future. I’m proud to say we have a great team where everyone has their area of focus and brings unique expertise to the table. Working together we and all our team push the boundaries of innovation and make AI-driven solutions a reality across the full range of industries.”
Favorite Memory
What has been your favorite memory working for the company so far? Abramov reflected:
“Our company had the privilege of contributing to a truly transformative project in the agricultural sector. The goal of this innovative initiative was to develop an AI-driven vehicle equipped with a laser system. It is capable of precisely targeting and eliminating weeds, as well as thinning crops — all without the use of harmful chemicals.”
“This technology represents a monumental achievement in sustainable farming. By avoiding chemical treatments, farmers can protect their crops without exposing them to potentially harmful substances. Beyond the crops themselves, this method has a profound positive impact on the surrounding ecosystem. Traditional herbicides and pesticides often harm wildlife, including small mammals and insects like bees.”
“Bees by the way play an irreplaceable role in pollination. However, their populations are in steep decline, with many species on the brink of extinction. Current methods of artificial pollination are still neither scalable nor effective, making the preservation of bee populations a top priority for humanity.”
“The laser-equipped robotics vehicle addresses this issue head-on. By targeting only weeds and harmful pests, the technology spares insects like bees from exposure to toxic chemicals. It provides a sustainable, efficient solution that promotes healthier ecosystems and more reliable crop yields.”
“Our role in this project was to annotate the vast amounts of visual data required to train the AI system. This involved creating precise labels for images of crops, weeds, and pests, enabling the AI to accurately distinguish between harmful and beneficial plants or organisms.”
“We are proud to have played a part in this project. It’s a great example of how AI and computer vision can drive real-world solutions to some of the most pressing challenges of our time. This technology is not just about farming efficiency; it’s about safeguarding the future of agriculture and protecting vital ecosystems.”
Core Products
What are the company’s core products and features? Abramov explained:
“Keymakr specializes in data annotation services, offering image and video annotation, data creation, data collection, and data validation for AI and machine learning (ML) computer vision projects of any size. With a robust technological infrastructure and deep expertise, we are equipped to handle data from diverse industries and domains.”
“Our motto, ‘Human teaching for machine learning,’ highlights our commitment to a human-in-the-loop approach. At the heart of our operations is an in-house team of over 600 skilled annotators. Our 4-level QA system combines human oversight with automation to ensure high-precision results.”
“We offer 24/7 service and tailor to meet each project’s unique needs.”
“To streamline and enhance the annotation process, we developed Keylabs.ai — our enterprise-grade annotation platform. Keylabs.ai supports all annotation types, offering a flexible and scalable solution for creating precise datasets tailored to each client’s needs.”
Challenges Faced
Have you faced any challenges in your sector of work recently? Abramov acknowledged:
“With AI systems influencing real-world decisions, the importance of ethical and unbiased data has grown. We carefully vet datasets to prevent biases that could propagate harmful outcomes in AI applications. This involves extensive checks to ensure fairness and inclusivity in the data we provide.”
“Another challenge is maintaining our ethical values while navigating market demands. For example, we’ve had to decline high-budget projects involving the annotation of explicit content. Despite the financial incentives, we prioritize the mental well-being of our operators and the promotion of responsible AI practices. We aim that our work aligns with ethical standards.”
“The rapid emergence of regulations, such as the European Union’s AI Act, poses one more challenge. Staying compliant with these evolving legal frameworks while meeting client demands requires agility and continuous adaptation.”
“The increasing reliance on automated annotation tools is a double-edged sword. While they offer efficiency, maintaining the quality, context, and ethical oversight that human annotators bring is crucial. Striking this balance is an ongoing challenge.”
“If we speak about current work challenges you can find more details about it here and here.”
“In such a fast-moving sector, staying ahead is essential. We view these challenges as opportunities to innovate and reaffirm our commitment to building responsible, ethical AI systems that can be trusted.”
Significant Milestones
What have been some of the company’s most significant milestones? Abramov cited:
“Since our founding in 2015 as a small startup, our company has achieved remarkable growth and established itself as an expert and professional in the data annotation market. By 2019, we had expanded to a team of over 600 employees — a number that continues to grow annually.”
“In 2020, we reached a significant milestone by serving over 100 enterprise clients and supporting thousands of commercial AI products with high-quality annotated data. Building on this momentum, we launched Keylabs.ai in 2021, an independent annotation platform designed to empower machine learning teams. Keylabs.ai handles all annotation types and easily integrates with ML-Ops solutions. This platform is now widely adopted by clients, competitors, and our internal teams, and it is continually enhanced with cutting-edge features to address evolving industry needs.”
“To date, we have sourced and created over 30 million images featuring various objects of interest and annotated more than 500 million objects, showcasing our commitment to delivering top-tier data solutions for AI innovation.”
Evolution Of The Company’s Technology
How has the company’s technology evolved since launching? Abramov noted:
“We have undergone significant evolution to meet the growing needs of the data annotation industry. Early on, we developed our proprietary annotation platform, and continually enhanced it.”
“We’ve introduced numerous features to automate repetitive tasks, streamlining workflows for our manual workforce. Additionally, we’ve prioritized the ergonomics of our tools to ensure convenience for both operators and clients.”
“Our recent release of Keylabs 2.0 represents a major leap forward, delivering a faster, smarter, and more efficient annotation process than ever.”
“We’ve also expanded the scope of our services. Today, we provide comprehensive video and image annotation, as well as custom data creation. If clients lack the data required for their models, we can create real, bespoke datasets tailored to their specific needs. This enables businesses to implement groundbreaking innovations with confidence.”
“Beyond technology, we’ve also focused on refining our internal processes and enhancing people and project management as the industry grows more complex. For example, we’ve implemented a robust learning and development program for our operators, offering regular training sessions and courses to ensure they remain at the forefront of the field.”
Customer Success Stories
Can you share any specific customer success stories? Abramov highlighted:
“We’ve achieved many successes, though the nature of the annotation business often requires confidentiality due to NDAs. However, we collaborate with leading players across industries such as automotive, manufacturing, agriculture, defense, and waste management. We take particular pride in the social impact of our annotation services.”
“One notable example is our recent partnership with WasteVision AI, a leader in waste management technology. By integrating Keymakr’s video and image annotation tools, WasteVision AI-enhanced its contamination detection system. This eliminated the need for drivers to manually check for contamination, significantly improving safety. Their computer vision-based system detects contamination and instantly notifies operators, improving the accuracy of waste sorting.”
“The impact of WasteVision AI’s technology has been remarkable, delivering haulers a 15-20x ROI while holding waste generators accountable. This innovation not only drives business growth but also promotes environmental sustainability. Keylabs platform further supports WasteVision AI by providing accurate and efficient annotation for their waste categorization efforts.”
“Of course, there are many more success stories that remain anonymous, particularly in the automotive and agriculture industries. One standout example includes a project involving “saving the bees,” showcasing the potential of annotation to address both industry and environmental challenges.”
Funding/Revenue
When asking Abramov about the company’s funding and revenue details, he revealed:
“We are a fully self-funded business, with our revenue doubling year-over-year since 2017. Although we faced significant challenges when COVID-19 struck in 2020, we managed to recover within a year.”
Total Addressable Market
What total addressable market (TAM) size is the company pursuing? Abramov assessed:
“The global artificial intelligence (AI) market size was valued at $515.31 billion in 2023 and is projected to grow from USD 621.19 billion in 2024 to USD 2,740.46 billion by 2032, exhibiting a CAGR of 20.4% during the forecast period. North America dominated the global AI market with a share of 41.23% in 2023.”
“We closely monitor market trends and respond to demand by advancing our technologies, enhancing the expertise of our annotators, and ensuring full compliance with all standards and regulations related to data usage.”
Differentiation From The Competition
What differentiates the company from its competition? Abramov affirmed:
“We are a boutique company recognized for the consistent quality and reliability of our outputs. Our in-house team includes experts from diverse domains — including automotive, agriculture, medicine, manufacturing, and more. It enables us to deliver precise, specialized solutions tailored to meet any industry need.”
“Many of our most satisfied clients initially worked with other providers before coming to us. This gave them a clear benchmark to appreciate our expertise and the difference we bring to the table.”
“While we may not always be the cheapest option upfront, we deliver long-term value by accelerating time-to-market and reducing risks. Partnering with a reliable provider like us isn’t just about minimizing the cost per task — it’s about ensuring your entire development pipeline stays on track.”
“Many companies put their best people on a pilot. So, you may get perfect datasets for your pilot, but then you’re locked into a contract with the provider and quality slips. It’s crucial to partner with providers who prioritize consistent quality and offer flexible terms.”
“We don’t structure contracts to tie you down. It’s in our best interest to consistently deliver top-quality data. If we fail to meet your expectations, you’re free to walk away and choose another provider — but we strive to ensure that never becomes necessary.”
Future Company Goals
What are some of the company’s future company goals? Abramov concluded:
“Our future goals are centered around contributing to impactful projects that benefit humanity. We aim to leverage our expertise to address critical global challenges, such as combating hunger, tackling environmental issues, and even advancing efforts to cure diseases like cancer. By participating in these transformative initiatives, we hope to make a meaningful difference in the world.”
