Abaka.AI has announced an $8 million fund to accelerate development and global rollout of the Abaka AI Data Platform, an end-to-end platform for cleaning, annotating, and supervising multimodal AI data. The capital will be used to expand production pipelines, add automation for complex supervision workflows like reinforcement learning from human feedback, and deepen model-in-the-loop evaluation. A portion of the funding will also support open benchmarks and interoperability work through the 2077AI Foundation, an open-source initiative co-founded by Abaka.AI to advance shared benchmarks and ethical data exchange, as well as promote interoperability across the AI ecosystem.
At the center of Abaka.AI’s rapid growth is Yunfei Zhao, co-founder and Chief Operating Officer. He has led one of the company’s largest commercial agreements to date with a leading AI company valued at $24B and grew MRR twentyfold year over year. Zhao co-founded the 2077AI Foundation to align capital and community with open research on data standards and, drawing on experience from Tesla and IDG Capital, is directing the new investment into greater production capacity and stronger evaluation for Abaka.AI’s enterprise customers.
Fixing AI’s Data Bottleneck
Not all data is made equal: most AI teams still struggle to source and maintain curated datasets that meet both product and compliance needs. The Abaka AI Data Platform addresses this with one platform that covers the full lifecycle from collection to annotation to delivery across text, image, audio, video, code, 3D/4D spatiotemporal data, and point clouds.
Abaka.AI’s multi-layer QA combines expert annotators, domain-specific training, consensus labeling, sampling audits, and automated error detection. Customers can deploy the Abaka AI Data Platform in the public cloud or on-premises, with options for hybrid environments with encrypted storage and role-based access control. For time-critical programs, AI-assisted pre-labeling plus human review increases throughput, boasting 24-hour turnarounds on urgent projects without trading away accuracy.
“The Abaka AI Data Platform is built to give enterprises of any size access to production-grade datasets and evaluation pipelines,” says Zhao. “We want to support the best possible AI outcomes, and those are only as strong as the datasets behind them.”
Built for Enterprise Production
For customers in highly regulated sectors like healthcare, finance, and embodied or autonomous systems, the platform’s configurable workflows and private deployment options help teams maintain compliance while scaling their data operations. Abaka.AI provides controls for the full data lifecycle, so teams can move from prototype assets to production baselines in one place. Evaluation tools and custom quality metrics map to each client’s acceptance criteria, which means models are tested under the same conditions they’ll face after launch.
“We understand the bar for rigor and privacy is extremely high,” Zhao says. “This funding lets us expand our platform and the services around it while pushing the industry toward open, high-quality practical standards.”
Leadership with Operator DNA
Zhao’s operator-investor background shapes Abaka.AI’s execution cadence and partnership strategy. Before Abaka.AI, he led global strategy and partnerships at Pika Labs, helping the company grow from seed to Series B and bringing the Pika 1.5 release into Hollywood production settings. That focus on shipping shows up in Abaka.AI’s momentum, with explosive revenue growth over three straight quarters and a distributed team in Silicon Valley, Paris, Tokyo, and Singapore. Zhao is directing hiring across production engineering and delivery and expanding the expert annotator network in specialized domains like STEM research and capital markets.
Interest across partners and the open-source community reflect Abaka.AI’s delivery record. Near-term priorities include more model-assist for complex labeling tasks, stronger on-prem and hybrid options with layered security and auditability, and continued support for the 2077AI Foundation’s benchmarks, interoperability work, and community programs. Zhao frames the roadmap as more automation where it’s safe, more controls where it matters, and wider coverage across data types and industries.
Looking Ahead
With fresh capital, Abaka.AI will expand their footprint among enterprises building next-generation conversational, reasoning, and vision systems. The company is investing in privacy-preserving workflows and richer domain coverage, so partners and clients can move from prototype to production with more support. Backed by $8 million and guided by operator-led execution, Abaka.AI aims to provide the dependable data infrastructure modern AI depends on, turning messy, multimodal inputs into repeatable, enterprise-ready pipelines.
