StartUp / Ai Startups

AI-native Data Infrastructure for Robotics

EnCord, co-founded by Eric and Ulrik, specializes in AI-native data infrastructure for physical AI and robotics, recently raising $60 million in Series C funding led by Wellington Management. The company emphasizes the critical role of data quality in AI model training, noting that even small dataset errors can greatly affect real-world model performance.
← back to ALL
yc_root_access • 2026-04-30T19:00:37Z
Source material: Robots Don't Need More Compute. They Need This.
Source material
Summary
EnCord, co-founded by Eric and Ulrik, specializes in AI-native data infrastructure for physical AI and robotics, recently raising $60 million in Series C funding led by Wellington Management. The company emphasizes the critical role of data quality in AI model training, noting that even small dataset errors can greatly affect real-world model performance. EnCord's platform acts as a universal data layer, allowing AI teams to efficiently create, manage, annotate, and evaluate data as AI models become increasingly complex. The founders identified the data challenges in AI development early in their careers, prompting them to establish EnCord prior to the surge in AI interest. The rise of ChatGPT has increased trust in AI, enabling companies to utilize AI for data processing that was once met with skepticism. EnCord is prioritizing multi-modal AI by integrating diverse data types, such as images, text, and audio, to improve physical AI applications. EnCord has established an R&D facility in the Bay Area to assist robotics companies in training their robots through dedicated data collection environments. The company aims to connect the real and digital worlds by providing services for both pre-deployment data collection and post-deployment support, including exception handling and observability.
Perspectives
EnCord's Approach
  • Prioritizes data quality to enhance AI model performance
  • Aims to be the leading platform for physical AI data
Challenges in Physical AI
  • Relies on human involvement for training due to data complexity
  • Faces competition and variability in data needs across applications
Neutral / Shared
  • Physical AI is expected to undergo a phase of consolidation
  • Timely decision-making is crucial for success in the AI sector
Metrics
60 million USD
Series C funding amount
This funding will support EnCord's growth and development in the AI sector
they just announced a 60 million Series C led by Willington Management
more than 300 units
of AI teams partnered with
A large customer base indicates strong market demand for their services
We work with more than 300 AI teams.
150 units
total workforce
A growing team reflects the company's expansion and operational capacity
We're a team of 150 people between London and San Francisco.
110 million USD
total funding raised to date
Total funding indicates the company's financial backing and investor confidence
we raised $110 million with the $60 million CBC that we just announced.
Key entities
Companies
EnCord • WeaverBuddix
Countries / Locations
ST
Themes
#ai_startups • #venture_capital • #ai_native • #ai_teams • #data_infrastructure • #encord • #physical_ai
Key developments
Phase 1
EnCord is focused on providing AI-native data infrastructure for physical AI and robotics, recently securing $60 million in Series C funding. The company emphasizes the critical role of data quality in AI model training, partnering with over 300 AI teams to enhance their data management processes.
  • EnCord, co-founded by Eric and Ulrik, specializes in AI-native data infrastructure for physical AI and robotics, recently raising $60 million in Series C funding led by Wellington Management
  • The company highlights the importance of data quality in AI model training, noting that even small dataset errors can greatly affect real-world model performance
  • EnCords platform acts as a universal data layer, allowing AI teams to efficiently create, manage, annotate, and evaluate data as AI models become increasingly complex
  • The founders identified the data challenges in AI development early in their careers, prompting them to establish EnCord prior to the surge in AI interest
  • EnCord currently partners with over 300 AI teams, including leaders in autonomous driving and robotics, and has expanded to a workforce of 150 across locations in London and San Francisco
Phase 2
Encord is developing AI-native data infrastructure specifically for physical AI and robotics, recently announcing a $60 million Series C funding round. The company is focused on enhancing data quality and management for AI applications, particularly in multi-modal contexts.
  • The rise of ChatGPT has increased trust in AI, enabling companies to utilize AI for data processing that was once met with skepticism
  • Encord is prioritizing multi-modal AI by integrating diverse data types, such as images, text, and audio, to improve physical AI applications
  • The company is tackling the complexities of physical AI, which requires real-world embodied data for effective data collection, unlike traditional language models
  • Encord has established an R&D facility in the Bay Area to assist robotics companies in training their robots through dedicated data collection environments
  • The company aims to connect the real and digital worlds by providing services for both pre-deployment data collection and post-deployment support, including exception handling and observability
Phase 3
Encord is developing AI-native data infrastructure for physical AI and robotics, recently securing $60 million in Series C funding. The company aims to enhance data quality and management for AI applications, particularly in multi-modal contexts.
  • Encords platform allows customers to efficiently index, curate, and annotate data, streamlining the model pipeline from pre-training to deployment, which accelerates market entry and revenue generation
  • The challenges of integrating multiple data modalities, such as video and audio, require significant human involvement in training, differing from the development of language models
  • Encord primarily serves robotics and self-driving car companies, helping them improve model performance and reduce time to market through its robust data infrastructure
  • WeaverBuddix, a notable customer, has successfully launched a laundry folding robot, demonstrating the practical applications of Encords physical AI data platform
  • Following a $60 million Series C funding round, Encord is poised to leverage the increasing interest in physical AI, which is a substantial economic opportunity as 80% of the global economy involves physical movement
Phase 4
Encord is developing AI-native data infrastructure for physical AI and robotics, aiming to enhance data quality and management for AI applications. The company recently secured $60 million in Series C funding to support its ambitious growth plans.
  • The trajectory of physical AI, like that of self-driving cars, is expected to go through initial hype followed by a phase of consolidation as companies refine their implementation strategies
  • Encord aspires to be the leading platform for physical AI data, aiming to collaborate with every physical AI company worldwide, similar to Stripes role in financial transactions
  • The company is enhancing its workforce by hiring and integrating AI agents, which improves both human and automated capabilities
  • A crucial lesson for founders is the necessity of making timely decisions, as indecision can lead to significant opportunity costs, highlighting the importance of agility in the AI sector
  • Founders should balance a long-term vision with flexibility, adapting to market signals instead of strictly adhering to a predetermined path

Related coverage

Closest startup themes

Related business and technology angles