Enterprise AI assistants are failing to deliver meaningful ROI because they cannot access current, trusted company data, forcing employees to manually reconcile conflicting sources and sometimes paste sensitive information into unapproved tools, raising compliance and exposure risks.
Datalab releases Omni Extract Bench, an Apache-2.0 open-source benchmark that evaluates structured-document extraction from Datalab, Reducto, Extend and LlamaExtract with normalized JSON scoring, Hungarian array matching and provider-paid API runs.
Reddit is ending RSS on November 13 and plans to shut public API access by March 2027 to curb AI-driven scraping, forcing researchers, social-listening tools and AI assistants to secure commercial data deals for programmatic access.
PageIndex SDK now lets developers index, retrieve from, and chat with text-based PDFs locally using their own LLM keys while retaining access to PageIndex Cloud through the same client, with PageIndex Flash becoming its default local tree-indexing method.
A local code-analysis system now pairs token search and graph expansion with optional LanceDB vectors, lifting reported recall from 58.2% to 71.4% while its SHA-256 ingest queue skips unchanged files and retries failures up to three times.
Google Research's TimesFM 3.0 has claimed the #1 ranking across all three major forecasting benchmarks, introducing powerful multivariate capabilities and zero-shot generalization, though its latest pretrained weights come with a non-commercial license restricting production use.
Databricks says its Genie Ontology returns business answers on day one, but that accuracy depends on the data foundation beneath it. In a September 1 guide, Srujan Alase and Richard Tomlinson lay out six layers, from entity resolution and Unity Catalog metadata to Metric Views and governance. The company recommends starting with one domain.
Cohere launches Parse, an enterprise document AI model priced at $1.50 per 1,000 pages — 85% cheaper than hyperscaler competitors — capable of processing complex tables, diagrams, and multilingual content across private cloud and on-premise environments, signaling a major industry shift toward affordable, domain-specific AI solutions.
Cohere launches Parse, a vision language model that converts complex enterprise documents into structured data, outperforming rivals like Mistral OCR 4, AWS Textract, and Google Document AI at just $1.50 per 1,000 pages, now available via the Cohere API, Microsoft Foundry, and AWS SageMaker.
Figure launches Index, the world's largest robot training dataset, boasting 264K app downloads, 16 million videos, and a $1 billion commitment to data and compute, aiming to power its Helix AI and deploy robots as an on-demand service.