←
SuperAnnotate451 posts
SuperAnnotate avatar✓

SuperAnnotate ✓

@superannotate

Platform, experts, and workflows to deliver the highest-quality AI data.

◷ Joined June 2019▧ 327 Photos and videos
Oscar Marz Örnberg avatar
⟳ SuperAnnotate retweeted
Oscar Marz Örnberg@omarzornberg·19h

We just put out a sample of CUA, the computer-use dataset I've been working on. → Real professionals doing their actual jobs → ~400 steps per task on average → 130+ apps, sheets to CAD and GIS → Task, screenshot, action and reasoning per step app.superannotate.com/public…

Post attachment
SuperAnnotate avatar
SuperAnnotate@superannotate·Nov 12

🚀 Introducing 𝗖𝗨𝗔: computer-use trajectories for the long, multi-app tasks where agents fail today. Agents handle short, well-defined computer tasks. The work that fills a professional’s day runs to hundreds of steps across several apps, and that’s where they loop, lose context and fail. CUA is recorded from professionals doing that work on their own computers: → 400+ steps per trajectory on average → 130+ apps, weighted toward visual work: video editing, 3D modeling, CAD, GIS → Task, screenshot, action and generated reasoning at every step → Natively video, usable for video-based training → SFT trajectories, with select tasks filtered toward RL tasks We made a sample of the dataset public. Click through the full trajectories here: app.superannotate.com/public…

Post attachment
SuperAnnotate avatar
SuperAnnotate@superannotate·11 Dec 2025

🎉Congratulations to our partner @databricks on the launch of the OfficeQA Benchmark. Enterprises can use the OfficeQA Benchmark to measure whether AI systems can handle the messy, high-precision tasks found in real business workflows. Teams can now more easily identify gaps, compare models, and make informed decisions about when AI is ready for deployment. The benchmark was developed using a large dataset: nearly 89,000 pages of historical U.S. Treasury Bulletins (documents spanning decades, with scanned pages, PDFs, complex tables, charts, figures, and mixed unstructured + structured data). 📣SuperAnnotate is proud to have powered the dataset and annotation rubrics behind this benchmark and to collaborate with the incredible Databricks team - Arnav Singhvi, Krista Opsahl-Ong, Jasmine Collins, @ivanzhouyq, @cindyxinyiwang, Ashutosh Baheti, Jacob Portes, Sam Havens, Erich Elsen, Michael Bendersky, @matei_zaharia, Xing Chen.

Post attachment
SuperAnnotate avatar
SuperAnnotate@superannotate·1 Dec 2025

📣 We’re at re:Invent this week and excited to welcome you at Booth 1022 together with NVIDIA. You can meet our founders, explore the latest developments, and learn how our partnership with NVIDIA shapes the next phase of enterprise AI. See you at the expo! #AWSreInvent

Post attachment
SuperAnnotate avatar
SuperAnnotate@superannotate·16 Oct 2025

POCs stall when teams can’t measure performance, leaving ML teams blind & leadership unsure. Our guide breaks it down: - Set the right metrics - Combine human + LLM review - Build reliable LLM judges Read: lnkd.in/eiHwE2xs

Post attachment
SuperAnnotate avatar
SuperAnnotate@superannotate·14 Oct 2025

🚀 Agent Hub just got an upgrade making it easier to use LLMs for data annotation & model evaluation. - Connect to models on Fireworks, Vertex, Databricks, Bedrock - Automate large-scale pre-labeling & evaluation - Enjoy faster, smoother workflows Read: buff.ly/7CgR7JH

Post attachment
SuperAnnotate avatar
SuperAnnotate@superannotate·8 Oct 2025

Need to ship better agentic, multimodal, and frontier AI faster and with high-quality? Join us: ⚡ London, Oct 16 📍 Databricks Data + AI World Tour | Booth K5

Post attachment