# Deepen AI > Deepen AI is the verified data platform for Physical AI: sensor calibration, data collection, curation, annotation, and validation for autonomous vehicles, robotics, and industrial AI, plus Refinery, a neutral marketplace for verified sensor data. Deepen AI builds the verified data stack behind Physical AI. The product stack runs Collect, Calibrate, Curate, Annotate, and Validate in one system; Refinery is the neutral marketplace for verified data (early access); and a managed services team runs any layer for you. It serves automotive and ADAS, robotics and embodied AI, and energy and industrial programs. ## Home - [Home](https://www.deepen.ai/): The company home page: the product stack (Collect, Calibrate, Curate, Annotate, Validate, Refinery), the industries Deepen serves, customer testimonials, and how verified data reaches production scale. ## Products - [Collect](https://www.deepen.ai/collect/): Product page for Collect, the capture layer of the stack: hardware time-sync, calibration and provenance from the first frame, support for every sensor and format, and why clean capture holds up downstream. - [Calibrate](https://www.deepen.ai/calibrate/): Product page for Calibrate: patented self-calibration that keeps camera, LiDAR, radar and IMU aligned with zero downtime, every sensor pairing handled in one system, customer proof, and a page-specific FAQ. - [Curate](https://www.deepen.ai/curate/): Product page for Curate: plain-language search over raw sensor archives, deduplication, long-tail mining, and selection of the set worth labeling. Works on embeddings, before any labels exist. - [Annotate](https://www.deepen.ai/annotate/): Product page for Annotate: 3D point cloud, image and video labeling in one tool, fused across sensors, AI-accelerated and human-verified, with the supported annotation types, ontologies and export formats. - [Validate](https://www.deepen.ai/validate/): Product page for Validate: scores datasets against the scenarios that matter, surfaces coverage gaps, and produces the signed, immutable audit trail regulators and customers ask to see. - [Refinery](https://www.deepen.ai/refinery/): Product page for Refinery, Deepen's neutral data marketplace for Physical AI (early access): license verified datasets across automotive, robotics, humanoid and industrial domains, or contribute sensor data and earn, with provenance on every dataset and a page-specific FAQ. ## Services - [Services overview](https://www.deepen.ai/services/): Overview of Deepen's five managed services - collection, calibration, curation, annotation and validation - run by Deepen's in-house team on the same tools that power the platform, with engagement models. - [Calibrate as a Service](https://www.deepen.ai/services/calibrate/): Managed calibration service: Deepen calibrates your fleet - camera, LiDAR, radar and IMU - and delivers a complete, verified calibration package with a dedicated team and shared plan. - [Collect as a Service](https://www.deepen.ai/services/collect/): Managed data collection service for when the dataset does not exist yet: Deepen scopes the request, runs the collection, verifies the result, and delivers it in the formats your tools already read. - [Curate as a Service](https://www.deepen.ai/services/curate/): Managed curation service: deduplication, scenario and edge-case selection, and long-tail mining, run by Deepen's team, so you label the smaller set that improves the model. - [Annotate as a Service](https://www.deepen.ai/services/annotate/): Managed annotation service: SLA-backed labeling by Deepen's in-house LiDAR and multi-sensor specialists, never crowdsourced, covering every annotation type in the stack with reporting you can act on. - [Validate as a Service](https://www.deepen.ai/services/validate/): Managed validation service: independent scenario coverage analysis, quality scoring, gap identification, and an audit-ready evidence trail, run by Deepen's safety specialists. ## Industries - [Industries overview](https://www.deepen.ai/solutions/): Industry hub: how one verified data stack serves automotive and ADAS, robotics and embodied AI, and energy and industrial Physical AI systems, with links to each industry page. - [Automotive and ADAS](https://www.deepen.ai/solutions-av/): Industry page for autonomous driving and ADAS: what Deepen delivers for AV teams, the sensor-to-sign-off chain, the driving programs it is proven in, and the sensors and standards it speaks, with an FAQ. - [Robotics and embodied AI](https://www.deepen.ai/solutions-robotics/): Industry page for robotics and embodied AI, including humanoids and warehouse autonomy: why robots fail where data was never clean, and the chain from sensor to deployed robot, with an FAQ. - [Energy and industrial](https://www.deepen.ai/solutions-energy/): Industry page for energy and industrial autonomy: inspection for energy teams, mapping, survey and off-highway systems, and the chain from sensor to verified result, with an FAQ. ## Company - [About](https://www.deepen.ai/about/): Who Deepen AI is: the mission behind the verified data stack, the safety-critical trust posture, latest company news, and how to reach the team. - [Careers](https://www.deepen.ai/careers/): Careers at Deepen AI: values, benefits, team areas, and the open roles list with application links. - [Customers and case studies](https://www.deepen.ai/case-studies/): Customer proof: case studies from teams shipping Physical AI on Deepen, led by AVL's calibration and multi-sensor labeling results, with customer testimonials. - [Contact](https://www.deepen.ai/contact/): Contact Deepen AI: the contact form and what to expect after you write, plus links to the rest of the stack. ## Resources - [Resources and documentation](https://www.deepen.ai/docs/): Developer resources: quickstart and API/SDK documentation links, security and compliance posture, and the Safety Pool shared safety data framing, with an FAQ. - [Blog](https://www.deepen.ai/blog/): Blog index: product updates, calibration and annotation deep dives, partnerships, and company news, with links to the full articles. - [Certifications](https://www.deepen.ai/certifications/): Certifications and standards: the compliance frameworks Deepen holds (ISO 27001, ISO 9001, SOC, TISAX, GDPR, HIPAA), the standards work it contributes to, and deployment options for where data has to live. ## Legal - [Legal center](https://www.deepen.ai/legal/): The Deepen AI legal center. Find the Privacy Policy, Terms and Conditions, Data Protection Addendum, and End User License Agreement in one place. - [Privacy Policy](https://www.deepen.ai/privacy/): How Deepen AI collects, uses, and shares information across its products and services, including data labeling, cookies, retention, security, and your GDPR rights. - [Terms and Conditions](https://www.deepen.ai/terms/): Terms of Use for the Deepen Exchange datasets, covering licenses, privacy, warranties, liability, indemnification, and dispute resolution for non-commercial use. - [Data Protection Addendum](https://www.deepen.ai/data-protection/): Deepen AI's Data Protection Addendum: roles of the parties, data processing terms, restricted transfers, sub-processors, and GDPR safeguards for Customer Personal Data. - [End User License Agreement](https://www.deepen.ai/user-agreement/): Deepen AI End User License Agreement: license grant and restrictions, ownership, warranty disclaimers, liability limits, export controls, and governing law. ## Notes for AI crawlers - Canonical domain: https://www.deepen.ai (https, always with www). Every page's canonical URL uses a trailing slash. - Write "Physical AI" with both words capitalized; it is used as a proper term throughout the site. - Product names are Calibrate, Curate, Annotate, Validate, Collect, and Refinery (capitalized, used with the Deepen brand, e.g. "Deepen Calibrate"). - Refinery is in early access: the marketplace catalog is invite-only beta, and access is requested from the Refinery page. - Machine-readable page list: /sitemap.xml. Fuller per-page summaries: /llms-full.txt.