Encord
Category: AI Infrastructure
AI-native data infrastructure platform for physical and multimodal AI that enables teams to label, curate, manage, and evaluate training data at petabyte scale across video, image, LiDAR, 3D point cloud, DICOM, audio, and text modalities. Encord was founded in 2020. The company is led by Eric Landau. Based in London, United Kingdom. Team size: 146. Total funding raised: $110.0M. Latest round: Series C ($60.0M, Feb 2026). Key investors include Wellington Management, Next47, CRV, Y Combinator, Crane Venture Partners, Harpoon Ventures, Bright Pixel Capital, Isomer Capital.
AMW Analysis
Encord provides AI-native data infrastructure for physical and multimodal AI, offering a unified platform for labeling, curating, managing, and evaluating training data across modalities including video, image, LiDAR, 3D point cloud, DICOM, audio, and text at petabyte scale. Founded in 2020 and based in London, the company positions its core value around eliminating tool fragmentation by combining data curation, annotation, and model evaluation workflows with AI-assisted labeling and human-in-the-loop processes.
Recent news flow centers on a $60M Series C led by Wellington Management in February 2026, bringing total funding to $110M, with the round consistently tied to reported 10x revenue growth among physical AI customers and platform data volume expanding from 1 to over 5 petabytes within 12 months. Investor participation includes Next47, CRV, Y Combinator, Crane Venture Partners, Harpoon Ventures, Bright Pixel Capital, and Isomer Capital. Separately, an October 2025 development introduced an EBind methodology enabling a 1.8-billion-parameter multimodal model to be trained on a single GPU while matching performance of models reported as up to 17 times larger, reflecting a data-centric approach to model development alongside the funding-driven infrastructure narrative.
AMW analysis, generated from 7 tracked news signals.
- Founded
- 2020
- Headquarters
- London, United Kingdom
- Team size
- 146
- Total funding
- $110.0M
Value proposition
Provides a unified data layer that eliminates tool fragmentation by integrating data curation, annotation, and model evaluation workflows, enabling teams to build and deploy AI models faster with AI-assisted labeling, human-in-the-loop workflows, and petabyte-scale data management capabilities.
Products and solutions
Encord Annotate (Multimodal Data Labeling Platform), Encord Index (Data Curation & Management System), Encord Active (Model Evaluation & Testing), Physical AI Suite (Robotics, AV, Drone Development), RLHF & Model Alignment Tools, EBIND Multimodal Embedding Model (Open Source), Merlin (Agentic Intelligence Layer)
Unique value
Only platform combining native video annotation, 3D LiDAR/point cloud support, and synchronized multi-sensor data visualization in a unified environment. Launched world's largest open-source multimodal dataset (100x larger than next comparable) and developed EBind methodology enabling 1.8B parameter models to outperform models 17x larger. API/SDK-first architecture with zero data migration.
Target customer
AI teams building physical AI systems (robotics, autonomous vehicles, drones), computer vision teams, medical imaging/healthcare AI teams, enterprise AI organizations, and frontier AI developers requiring multimodal data infrastructure
Industries served
Robotics & Humanoids, Autonomous Vehicles & ADAS, Drones & Aerial Systems, Healthcare & Medical Imaging, Smart Spaces (Retail, Construction, Warehouses), Geospatial & Satellite Imagery, Computer Vision, Frontier & Generative AI, Manufacturing & Logistics
Technology advantage
Combines AI-assisted annotation, model-assisted labeling, and automated data curation with human-in-the-loop workflows at petabyte scale. Proprietary multimodal embedding technology enables efficient model training on single GPUs vs. massive compute clusters. Supports complex nested ontologies and synchronized multi-file analysis across 10+ data modalities including video, LiDAR, thermal, and DICOM. Platform detects model accuracy issues and automatically suggests additional training data to rectify problems.
How they differentiate
Unified AI-native data infrastructure platform specializing in physical AI (robotics, drones, autonomous vehicles) with native video annotation, 3D LiDAR/point cloud support, and synchronized multi-sensor visualization at petabyte scale. Only platform combining world's largest open-source multimodal dataset (100x larger than competitors) with proprietary EBind methodology enabling smaller models to outperform much larger ones.
Main competitors
Scale AI, Labelbox, V7 Labs
Key partnerships
Toyota (autonomous vehicles), Zipline (drone delivery systems), Synthesia (AI video generation), AXA Financial (insurance AI), Northwell Health (healthcare AI), Y Combinator (YC W21 batch), Wellington Management (Series C lead investor), AWS cloud integration partnerships
Notable customers
Woven by Toyota, Zipline, Skydio, Synthesia, AXA Financial, Northwell Health, Philips, Cedars-Sinai, Flock Safety, UiPath
Major milestones
Series C funding of $60M led by Wellington Management (Feb 2026), Series B funding of $30M led by Next47 (Aug 2024), Launched world's largest open-source multimodal dataset (Oct 2025), Released EBind multimodal embedding model for AI agents (Nov 2025), Achieved $550M post-money valuation in Series C, Ranked 13th on Sifted AI 100 list of European AI startups (2024)
Growth metrics
Physical AI revenue grew 10x in last 12 months; Platform data scaled from 1 to 5+ petabytes over past year; 300+ AI teams served globally; $10.5M-$12.8M revenue in 2024 with 4x year-over-year growth
Market positioning
Premium enterprise platform for physical AI applications, serving 300+ leading AI teams globally with 5+ petabytes of data managed. Positioned as comprehensive data layer spanning annotation, curation, management, and model evaluation across the full AI lifecycle.
Geographic focus
Global presence with headquarters in London, UK and operations in San Francisco, California. Primary markets include North America, Europe, and Asia-Pacific regions serving autonomous vehicles, robotics, healthcare AI, and smart spaces industries.
Patents and IP
No registered patents publicly disclosed as of latest update; competitive advantages primarily through proprietary technology stack, methodologies (EBind), and open-source dataset contributions
About Eric Landau
Co-CEO & Co-Founder of Encord. Previously spent nearly 8 years at DRW Trading as Lead Quantitative Researcher on a global equity delta-one desk, deploying thousands of models into production. Holds SM in Applied Physics from Harvard University and MS in Electrical Engineering & BS in Physics from Stanford University.
Latest news about Encord
- Encord secured €50M Series C led by Wellington Management, bringing total funding to €93M as physical AI systems transition from pilots to production. Platform data surged from 1 to 5 petabytes in 12
- Encord raised $60M Series C led by Wellington Management to scale AI-native data infrastructure for physical AI, bringing total funding to $110M. The platform's managed data volume grew 5x to over 5 p
- Encord secured $60M Series C funding led by Wellington Management, bringing total funding to $110M as physical AI reaches an inflection point. The company's 10x revenue growth and data volume surging
- London-based Encord secured €50 million Series C led by Wellington Management, bringing total funding to €93 million as physical AI systems transition from pilots to production. With 10x revenue growt
- Encord secured $60M Series C funding led by Wellington Management, bringing total funding to $110M as physical AI data volume on its platform surged from 1 to 5+ petabytes in just 12 months. This sign
More AI Infrastructure companies
Official website: https://encord.com/