Realtime AI News
Sqwish Labs offers decision-model customization trained on customer data
Sqwish Labs is offering customization for decision models, letting customers train the system on their own data to extend what it can do, TechCrunch reported on November 17. The launch is framed against a backdrop of rapidly rising decision-model adoption.

Sqwish Labs is offering customization for decision models, letting customers train the system on their own data to extend what it can do, TechCrunch reported on November 17.
The story frames the move with a question in its headline — as decision-model use skyrockets, is customization the next frontier — and presents Sqwish Labs as a new wrinkle on the category.
A decision model here means one that returns a recommendation, ranking or trade-off given business context, rather than acting as a general-purpose text assistant. The two are judged differently: one on whether its judgement matches the business rule, the other on whether the output is fluent and accurate.
Training on a customer's own data is meant to close the gap between a generic model's output and the judgement a specific workflow actually applies, which typically needs manual correction before it is usable.
Customization also carries a cost. Once customer data enters the training process, data governance, access control and responsibility for model updates all have to be redefined, which is a real hurdle in procurement.
The available summary does not disclose Sqwish Labs' training approach, the deployment options it supports, pricing, or named customers, so the announcement reads more as a statement of direction than as an evaluable product.
Three things to watch: whether customer case studies and outcome metrics appear, how far the cost of a customized model diverges from a general one, and how the promise of training on your own data survives compliance review.
Why it matters
Competition in decision models is shifting from raw capability toward how closely a model matches a customer's own judgement, and customization is the shortest route there. The harder negotiation starts afterwards, when data governance and responsibility for model updates have to be settled.
Nearby Updates
All11/18, 02:04
Pedestal AI raises $4.7M to put agents to work inside supply chains
Pedestal AI has raised $4.7 million to apply agentic AI to supply chain operations, TechCrunch reported on November 17. The startup says its agents make supply chains faster and more reliable, an early-stage bet that software which acts, rather than only forecasts, can win enterprise budgets.
11/18, 01:57
Anthropic opens linked billing beta so Claude API apps can charge their own customers
Anthropic launched linked billing in beta on November 17, letting developers charge Claude API usage in their applications to a customer's own Claude organization. Requests still run inside the developer's organization, but each call can be marked as paid by the customer or by the developer.
11/18, 01:45
OpenAI launches GPT-6.2 Astra and Sol for professional work, coding and computer use
OpenAI introduced GPT-6.2 Astra and Sol on November 17, describing better professional outputs, coding, computer use and efficiency at lower cost across ChatGPT Work, Codex and the API. Both models arrive in one announcement without a stated division of labour, so the practical differences will only surface in documentation and pricing.
11/18, 01:32
Proton brings Lumo to the desktop with per-connector permission controls
Proton launched Lumo Desktop on November 17 for Windows, macOS and Linux, letting its AI work with local files and connected apps to automate tasks. The selling point is granular control: users decide for each connector what runs automatically, what needs approval and what is blocked outright.