As of November 2026, Latka tracks 18 deep learning software companies with $5M–$10M in annual revenue. They have combined revenues of $118.5M and employ 892 people. They have raised $398.7M.
Every company below sells deep learning software to other businesses and is ranked by its most recent annual revenue. Revenue, funding, headcount and customer figures come from CEO interviews on the Latka podcast, public company filings, and Latka estimates where a company has not disclosed a number.
Deep Learning Software refers to a category of applications that leverage deep learning techniques to analyze data, automate processes, and derive insights. These tools use artificial neural networks to mimic human cognitive processes, allowing for complex computation and pattern recognition. Common use cases include image and speech recognition, natural language processing, and predictive analytics across various industries such as healthcare, finance, and technology. Typical features of deep learning software include data preprocessing tools, model training capabilities, performance optimization, and deployment functions. Users often span a wide range of professions, including data scientists, IT professionals, and analysts, who apply these technologies to derive actionable insights from large datasets. As the technology matures, industries are increasingly adopting these solutions to enhance decision-making processes and drive innovation in their operations.
Sorting: Highest -> Lowest
Showing 10 of 18 companies ranked by annual revenue.
Karlsruhe, Germany
We believe that AI is the most powerful tool our generation has at hand. At understand. ai , we make deep learning accessible for real-world applications like self-driving cars. More Info: https://understand. ai /privacy-policy Founded in 2017 in Karlsruhe, Germany, we are the ideal annotation partner throughout...
Mountain View, California, United States
Provides GPU cloud solutions for AI training, inference, and deployment; focuses on high-performance infrastructure at scale.
Amsterdam, North Holland, Netherlands
Quatt is an ambitious start-up that was founded in 2021 by two entrepreneurs and brothers from the Netherlands. Our goal is to accelerate the transition to sustainable heating in the Netherlands and abroad. Our first product is a smart, hybrid heat pump for the Dutch market. Coming winter we will launch our second product: the All-Electric upgrade, enabling consumers to permanently say goodbye to gas, completely at their own pace. The Dutch government is strongly incentivizing 8 million homeowners to get rid of natural gas and start heating sustainably within the next few decades. Unfortunately, the current pace of heat pump installations is far too low. Therefore, Quatt is improving the overall customer experience, adding smart control software, simplifying the installation job and, last but not least, greatly enhancing the aesthetics. By doing so, we want to increase adoption by an order of magnitude. We have developed a highly advanced control system based on several AI technologies, like machine learning and deep learning. This system enables us to operate our heat pumps, in conjunction with already installed natural gas heaters, in the most cost- or energy-efficient way, at any given day. This can reduce gas by up to 80%. By combining this with D2C sales & installation, Quatt is aiming to reach a payback time of 3-4 years; and thus a very quick uplift in sales. We are working towards an installed base of 3 million heat pumps.
Cambridge, England, United Kingdom
Neural interfaces to understand and use nerve data in medicine.
Berlin, Berlin, Germany
Developer of robot control system designed for complex assembly tasks. The company's robots perform movements that react to minute changes in the environment, as perceived by the robot with cameras or force sensors, and are trained through demonstration, correction and repetition, enabling users to create skills without writing a single line of code or modelling the problem.
Lyon, Auvergne-Rhône-Alpes, France
Operating in 13 European countries, ProovStation is a DeepTech start-up specializing in computer vision. Since 2017, its goal has been to scale up, automate and standardize the automotive inspection process worldwide to automatically estimate repair costs. Its mission is to reduce the total cost of vehicle inspection and increase the efficiency of the entire process to improve the customer and employee experience. ProovStation develops a range of high-tech vehicle scanners which, assisted by a combined array of optical technologies, conduct a 360° scan of the vehicle in just three seconds. This technique covers all angles and analyzes all damages at millimeter scale. The “Station” uses artificial intelligence to detect, measure and locate all of the damages sustained by the vehicle. It then generates a time-stamped, geolocated and secured report. ProovStation customers are subsequently given access to range of SaaS solutions (automatic reconditioning quotes; automatic vehicle pricing for resale; check in/check out, vehicle identification; 2D/3D photo studio; and activity management).
Menlo Park, United States
Periodic Labs uses AI to analyze large-scale scientific data and collaborates with academic and industry partners to support material science research. It aims to build AI that can accelerate discoveries in physics, chemistry and other fields.
Buffalo, New York, United States
Developer of a technology platform designed to revolutionize the diagnosis, monitoring and treatment of brain-related disorders. The company's technology platform combines big data analytics, artificial intelligence, and brain network activation mapping to quantify function and dysfunction, disease progression, and therapeutic effects, enabling businesses to get elucidation of the way the brain works at high temporal and spatial resolutions.
Latka tracks 18 deep learning software companies with $5M–$10M in annual revenue. Together they generate $118.5M in annual revenue and employ 892 people.
understand.ai is the largest, with $8.9M in annual revenue, founded in 2017.
The average deep learning software company in this list makes $6.6M a year, across 18 companies with reported revenue.
Ranked by annual revenue, the leaders are understand.ai, Verihubs, GMI Cloud, Quatt and Atomwise.
The 18 deep learning software companies with $5M–$10M in annual revenue tracked here have raised $398.7M in disclosed funding between them.
- Must provide tools for building, training, and deploying deep learning models - Should include support for various data types, such as text, images, and audio - Must offer capabilities for model evaluation and performance metrics - Should enable integration with big data frameworks and cloud services - Not just focused on traditional machine learning; must specifically address deep learning methods - Must facilitate automation of repetitive tasks within the deep learning workflow - Should be suitable for use by professionals such as data scientists, engineers, and researchers
