Data Labeling Startups

Spade provides transaction data enrichment and standardization for financial institutions, fintechs, platforms, and AI companies. The company transforms inconsistent transaction streams into structured, geolocation-matched intelligence with fast processing and high accuracy.

6 investors

Sepal AI generates training and evaluation datasets for AI models at scale, working with leading AI research labs and enterprises to improve model performance on real-world tasks.

5 investors

Sieve creates high-quality multimodal datasets and training environments for frontier AI labs developing generative media, robotics, and agentic systems. The company combines exabyte-scale infrastructure with novel data collection and processing techniques to serve AI research teams and enterprise customers.

4 investors

AfterQuery is an applied research lab that curates training data solutions for foundation model development by capturing expert reasoning and decision-making processes. The company works with domain experts to transform real-world professional work into structured datasets that teach AI models how to solve complex problems through step-by-step reasoning rather than just outputs.

3 investors

BeSimple collects, licenses, and annotates high-quality conversational audio datasets across 15+ languages for training and evaluating audio and multi-modal AI models. The company serves AI teams building speech recognition, voice, and conversational AI applications.

2 investors

Centaur.ai uses performance-based AI and collective intelligence methods to improve medical decision-making in skin lesion assessment for cancer detection. The company combines expert human judgment with AI algorithms weighted by demonstrated accuracy rather than credentials.

2 investors

Deasy Labs automates the transformation of unstructured data into AI-ready structured data through a context engine that handles discovery, tagging, filtering, and enrichment. The platform serves engineers and data scientists building AI projects that require clean, continuously maintained data to prevent model drift in production.

2 investors

JumpWire provides data access control software that sits between applications and databases to identify and restrict access to sensitive fields based on organizational policies. The platform enables fine-grained permission controls for who can read or modify specific data properties, and works within existing infrastructure without requiring data migrations or application changes.

2 investors

Lightly provides tools for AI teams to manage and curate training data more efficiently. The company focuses on data-centric approaches to improve machine learning model quality.

2 investors

Unbound builds security infrastructure to govern AI coding agents in enterprise environments. The company provides an Agent Access Security Broker that sits between AI agents and systems they interact with, enabling security teams to discover, audit, and control agent actions without disrupting developer workflows.

2 investors

Cartpole develops simulation environments designed to train advanced machine learning models using reinforcement learning techniques.

1 investor

Coasty provides AI agents that automate healthcare revenue cycle tasks like eligibility verification, prior authorization, claim submission, and denial management by performing repetitive work across EHRs, practice management systems, and payer portals through screen interaction. The platform is designed for healthcare billing teams and operates within customer environments with full audit trails and approval gates for payer-facing submissions.

1 investor

Datoric sources and licenses ethically obtained voice, video, and multimodal training data for frontier AI development, working directly with consented, compensated contributors such as native speakers and domain experts. Each dataset carries verifiable provenance and clean licenses aimed at AI model teams building multimodal systems.

1 investor

Instance Labs provides verification software that evaluates whether robots have successfully completed assigned tasks from camera footage. The platform helps robotics companies validate training data and accelerate policy development by analyzing raw video episodes and returning success verdicts with evidence.

1 investor

Liva AI collects specialized training data for audio and video models used by AI research labs, addressing gaps in publicly available datasets. The company serves machine learning teams developing speech and video recognition systems.

1 investor

Markov Studios captures screen recordings and interaction data from real computer use to create datasets that machine learning models can learn from. The company produces labeled data from real workflows in productivity software and video games.

1 investor

Sciloop builds datasets and solutions for frontier AI models by connecting a global network of expert mathematicians, physicists, computer scientists, and domain specialists to solve complex problems that current AI cannot address.

1 investor

A video dataset company that provides cleaned, segmented, and labeled video content from millions of hours of footage for AI model training. The platform serves AI research labs and development teams.

1 investor

A data collection company that gathers high-quality human data across diverse interactions and environments to support development of multimodal and embodied AI systems.

1 investor