Claude turns live data into inspectable dashboards
Live data, low-latency agents, and a useful data-fusion recipe make today’s list.
In today’s edition
- Claude turns connected data into live dashboards.
- OpenAI speeds up GPT-6.1 Sol for live work.
- Claude Science maps a data-fusion workflow.
- Anthropic commits $150 million to federal science.
Claude turns connected data into live dashboards
Claude Dashboards is a paid-plan beta for asking questions of a connected data platform or CRM in plain language. It writes the query, builds the chart, refreshes it as underlying data changes, and exposes the query behind every chart.
- Connect enterprise sources including Snowflake, Databricks, Amazon Redshift, ClickHouse, and existing connectors such as Salesforce.
- Click a dashboard number to inspect the query that produced it, or ask Claude to explain the result.
- Send a dashboard to tools such as Grafana, Hex, Mixpanel, PostHog, or Sigma for deeper analysis.
- Enterprise admins must enable Dashboards, and viewers run queries through their own connections by default.
What it means Use this to prototype self-serve analytics where each chart stays auditable, rather than treating natural-language dashboards as a replacement for BI governance.
Source: claudeai
GPT-6.1 Sol gets an Ultrafast lane
OpenAI is rolling out an Ultrafast service tier for GPT-6.1 Sol across the API, Codex, and ChatGPT Work. OpenAI says it delivers up to 8× the speed of Sol Standard while retaining near-Astra intelligence.
- API pricing is $12 per million input tokens and $60 per million output tokens.
- OpenAI positions the tier for outage debugging, agents navigating apps, and live experiences where seconds matter.
- The tier is available in supported regions, including US and EU data residency support.
- Codex and ChatGPT Work access is limited to Pro 500, eligible usage-based Enterprise, and credit-based Edu plans; Enterprise admins must enable it.
What it means Reserve this tier for live agent workflows where waiting is expensive, and price the latency gain against its $12 input and $60 output token rates.
Source: OpenAIDevs
A repeatable recipe for agent-led data fusion
An astrophysicist used Claude Science to combine scattered ultraviolet sky surveys into a complete map, including estimates for the roughly one-third of the sky never observed in UV. The detailed account is also a strong pattern for messy, multi-source data projects.
- Agents located public surveys, downloaded the data, corrected inconsistencies, and cross-calibrated instruments onto a shared coordinate system.
- For missing UV regions, the system learned relationships between mapped UV data and visible, infrared, and radio observations.
- The team tested predictions by hiding known UV regions, then reported estimates within about 10% of the withheld measurements.
- Human review still mattered: an astrophysicist spotted artifacts that two agent review rounds missed, then directed a correction across 38,000 observations.
What it means Borrow the pattern for messy multimodal data: use held-out checks and uncertainty labels, then keep a human visual QA pass before trusting agent-produced output.
Source: www.anthropic.com
Anthropic pledges $150M for US federal science
Anthropic has committed $150 million over three years to the Genesis Mission, a federal effort focused on using AI to accelerate scientific and technological discovery. The company says the funding will make Claude available to more than 15 agencies, including NASA, NIH, and NSF.
- The commitment runs for three years and is a funding pledge, not a new government policy or rule.
- Anthropic says it has already worked with the Department of Energy to bring Claude to scientists across national laboratories.
- The current commitment expands support to agencies participating in the Genesis Mission.
- Anthropic announced the pledge at a White House science summit.
What it means This is a material lab-to-government commitment that makes scientific-agency adoption a competitive front, even though the three-year pledge is not a new public policy.
Source: www.anthropic.com