Encord , an AI data development platform for advanced vision and multimodal teams, announced a $30 million Series B funding round. The round was led by Next47, a global venture firm that invests in early and expansion stage companies in the SaaS, AI and enterprise sectors, with participation from existing investors Y Combinator, CRV and Crane Venture Partners.
The effectiveness of an AI model is directly tied to the quality of the data it is built on. And as executives face increasing pressure to integrate AI into their enterprises, they are often faced with the challenge of siloed, uncurated, and unprepared data, severely hindering the model development process. This problem is a major bottleneck preventing widespread adoption of AI. Moreover, the expansion of AI has highlighted another critical problem: a shortage of high-quality data and the human domain experts needed to prepare this data, evaluate models, and provide actionable AI training and fine-tuning.
Encord leverages automation rather than relying on human oversight to give AI teams everything they need to AI-ready large datasets and put models into production faster. Encord’s software addresses all four steps of transforming data into AI: data management, curation, annotation, and model evaluation.
By consolidating this workflow into one platform, companies have a trail of evidence that shows why models make certain decisions, making it easier for companies adopting AI to uncover weak spots and identify the correct data needed to retrain models for better results. With a software-first approach, Encord can easily scale as customers’ data volumes and model outputs grow, addressing the limitations of other tools and human oversight.
Encord also launched Encord Index, an end-to-end data management tool that enables customers to visualize, search, sort, and control the in-house data they use to train and create AI models. Index seamlessly integrates with data storage, including AWS S3, GCP Cloud Storage, and Azure Blob, and automates the process of curating the best data and removing useless or biased data. As a result, Encord customers have reduced dataset size by 35% by curating the best data, improved model performance by over 20%, and saved hundreds of thousands of dollars in compute and human annotation costs.
Encord’s platform is built on vision, an inherently more complex modality than text, positioning the company to lead in the impending multimodal era and ahead of LLM-only focused companies in the current hype cycle. Encord also provides data management for AI models based on images, videos, medical images, visual language, speech, and more. With over 2,000 predictive and generative models in production, the company is trusted by over 120 top AI teams at organizations such as Synthesia, Philips, Zoopla, Cedars-Sinai, and Northwell Health.
Important Quotes:
“In 10 years, every company will have an AI department just like they have an IT department today, or risk becoming irrelevant. Encord will be the only platform you need to accomplish the process of transforming data into functional AI models. As AI models tackle increasingly complex problems, accessing and managing data will become increasingly important. By leading with automation, we’ve built a scalable solution to support our customers at every stage of their AI journey.”
- Ulrik Stig Hansen, Co-founder and President of Encord
“State-of-the-art models like our recently released Expressive Avatar foundation model, EXPRESS-1, require extremely sophisticated infrastructure to succeed. Encord Index is a high-performance system for our AI data that can be sorted and searched at any level of complexity, supporting our ongoing commitment to pushing the boundaries of AI avatar technology to meet our customers’ needs.”
- Victor Liparberg, co-founder and CEO of billion-dollar generative AI company Synthesia
“We’re seeing a lot of innovation at the model and compute layers of AI, but not as much at the data layer. Encord’s leading platform tackles one of the biggest challenges in AI today – understanding and managing the data that will drive high-quality, reliable results from enterprise AI applications, thereby reducing the risk of AI implementations. The market has long needed a solution to the AI data problem, and we’re proud to back Encord as the startup that has found it.”
- J. Rylander, General Partner, Next47
