# agentsite-static-version=144d43e7-3731-4867-8e6c-02e143ad1d86 # unsupervised.com > AI-optimized mirror of unsupervised.com containing 48 pages totalling 31,393 words of clean markdown content, structured data, and semantic HTML. Original source: https://unsupervised.com. Last updated: 2026-10-05T16:44:42.432Z. Each page is available as HTML (with JSON-LD structured data) and Markdown (text-only, ideal for LLMs and RAG). ## Homepage - [Unsupervised — Intelligence Is No Longer the Bottleneck](/content/site-root.html): What AI is worth will be set by how much work it can do without a person watching. Unsupervised builds the systems that make that trust rational. (457 words) ## Articles & Blog Posts - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/terms/index.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (5,020 words) - [What Is Driving the Increase in Medicaid Spending?](/content/examples/hhs-medicaid/report-html.html) (1,372 words) - [pdf-materials/unsupervised-20retailer-20supply-20chain-20case-20study-pdf.html](/content/pdf-materials/unsupervised-20retailer-20supply-20chain-20case-20study-pdf.html) (416 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/ethics/index.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (625 words) - [pdf-materials/unsupervised-20ecomm-20case-20study-pdf.html](/content/pdf-materials/unsupervised-20ecomm-20case-20study-pdf.html) (1,044 words) - [AI Agents for Retail - Unsupervised | Unsupervised](/content/solutions/retail/index.html): Transform retail data into actionable insights. Optimize promotions, understand customers, and predict demand with AI. (558 words) - [HHS Medicaid Run — Published Example | Unsupervised](/content/examples/hhs-medicaid/index.html): A full overnight Finder run over 194,002,693 rows of Medicaid fee-for-service claims, from one open-ended prompt to a source-backed HTML report. (1,367 words) - [Structure Makes Frontier Models More Expensive and Open-Weights Models Cheaper | Unsupervised](/content/resources/what-agent-work-costs/index.html): We ran the same 104-task agent benchmark across a frontier model and an open-weights model, raw and under structured workflows, then went back to the logs for the bill. Frontier dollars are billed; open-weights dollars are measured tokens times a recorded rate. Structure nearly tripled the frontier cost per task and made the open-weights run cheaper. The exact figures, with the work shown. (1,304 words) - [pdf-materials/unsupervised-20call-20center-20case-20study-pdf.html](/content/pdf-materials/unsupervised-20call-20center-20case-20study-pdf.html) (219 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/resources/two-50m-milestones-announcing-our-series-b-and-over-50m-found-by-customers.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (362 words) - [resources/index.html](/content/resources/index.html) (211 words) - [Workflows Written by a Frontier Model Made an Open Model Twice as Fast | Unsupervised](/content/resources/open-weights-borrowed-workflows/index.html): We ran an open-weights model over 104 real one-day agent tasks two ways: raw, and executing workflows a frontier model had written. The borrowed workflows finished in under half the wall-clock time. The speed result is solid; the score edge needs more runs. An early readout, with the work shown. (990 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/resources/unsupervised-s-ai-analysts-cross-1-billion-found/index.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (908 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/resources/pinpointing-kpi-changes-in-minutes-using-unsupervised.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (608 words) - [Unsupervised Named in Gartner's Market Guide for Agentic Analytics for the Second Consecutive Year | Unsupervised](/content/resources/gartner-2026-market-guide-agentic-analytics/index.html): Gartner published its second Market Guide for Agentic Analytics in February 2026, naming Unsupervised a Representative Vendor in both editions. The report predicts that by 2028, 60% of agentic analytics projects relying solely on MCP will fail without a consistent semantic layer. (933 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/resources/security-and-privacy-at-unsupervised/index.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (808 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/resources/finding-opportunities-and-building-action-plans-with-unsupervised-ai-analyst.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (622 words) - [AI Agents for Marketing - Unsupervised | Unsupervised](/content/solutions/marketing/index.html): Achieve 10X ROI with AI-powered marketing analytics. Optimize campaigns, understand customers, and predict outcomes. (556 words) - [AI Agents for Health Insurance Payers - Unsupervised | Unsupervised](/content/solutions/health-insurance/index.html): Transform healthcare data into strategic insights. Find $58M+ in opportunities with AI-powered analytics for health insurance payers. (726 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/resources/avoiding-the-risks-of-llm-hallucinations-in-sql-generation.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (626 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/resources/using-unsupervised-to-expand-your-impact-and-target-the-right-audience.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (470 words) - [AI Agents for eCommerce - Unsupervised | Unsupervised](/content/solutions/e-commerce/index.html): Increase sales by 400% with AI-powered ecommerce analytics. Optimize conversions, personalize experiences, and predict trends. (597 words) - [AI Agents for Financial Services - Unsupervised | Unsupervised](/content/solutions/finance/index.html): Achieve 5X ROI with AI-powered analytics for financial services. Automate risk analysis, fraud detection, and customer insights. (519 words) - [AI Agents for Technology - Unsupervised | Unsupervised](/content/solutions/tech/index.html): Get 8.2X ROI with AI analytics for technology companies. Understand user behavior, optimize products, and predict trends. (596 words) - [Introducing DeepWork: Structured Workflow Automation | Unsupervised](/content/resources/introducing-deepwork-structured-workflow-automation.html): We're releasing DeepWork, a source-available workflow framework that adds structure and quality control to AI-assisted tasks. (497 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/resources/simplify-data-queries-with-ai-agents-that-do-data-analysis.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (602 words) - [AI Agents for Telecom - Unsupervised | Unsupervised](/content/solutions/telecom/index.html): Discover $180M+ in opportunities with AI analytics for telecom. Reduce churn, optimize truck rolls, and improve customer satisfaction. (514 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/resources/dtc-retailer-uncovers-2-million-in-new-opportunities-with-unsupervised.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (527 words) - [Writing | Unsupervised](/content/writing/index.html): Research, field notes, and product updates from our work with agents. (211 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/resources/capture-value-in-your-text-fields-with-ai-questions.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (640 words) - [2119 — Tests Your Agent Can’t Game | Unsupervised](/content/2119/index.html): Spec-driven test enforcement for coding agents: plans become RFC 2119 requirements, every MUST gets a test, and a fresh-context reviewer judges whether each test would fail if the requirement broke. (526 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/resources/fortune-500-retailer-finds-1m-with-improvements-to-supply-chain-management.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (428 words) - [Panopticon — A Fleet of Agents. One Operator. | Unsupervised](/content/panopticon/index.html): Panopticon runs a fleet of coding agents in parallel (each in its own container, on its own branch) and surfaces only the tasks waiting on your judgment. (496 words) - [Products | Unsupervised](/content/products/index.html): Agent control systems and agentic data analysis: tools for managing agents and trusting their output. (173 words) - [Fortune 500 Healthcare Payer Finds $58 Million in Member Growth and Retention Opportunities | Unsupervised](/content/resources/fortune-500-healthcare-payer-finds-58-million-in-member-growth-and-retention-opportunities.html): A Fortune 500 healthcare payer used Unsupervised to merge and analyze 26 data sets, finding $58M in member growth and retention opportunities in the first six months across 60 million beneficiaries. (369 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/resources/unsupervised-customers-have-now-found-over-300-million-using-our-ai.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (278 words) - [DeepWork — Do It Once. Run It Forever. | Unsupervised](/content/deepwork/index.html): DeepWork turns a task you have done once with your agent into a repeatable, quality-gated workflow. Workflows, reviews, and schemas that enforce the work. (321 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/databricks/index.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (380 words) - [Finder — Evidence Before Action | Unsupervised](/content/finder/index.html): Finder searches vastly more of your data than query-by-query exploration and hands your agent ranked evidence, caveats, and open questions. (265 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/privacy/index.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (1,882 words) - [AT&T puts $100M+ of insights into action | Unsupervised](/content/resources/at-t-finds-100m-using-unsupervised-data-agents/index.html): AT&T has put Unsupervised insights into action across its business units, with an estimated value of more than $100M, and is expanding its use of our AI Data Analysts over the next year. (178 words) - [Outfitter — Dotfiles for Your Coding Agent | Unsupervised](/content/outfitter/index.html): Outfitter launches your agent CLI from profiles that compose context, prompts, models, skills, extensions, and workflows: reproducible, shareable, versioned. (240 words) - [Unsupervised CLI — Install the Analysis Launcher | Unsupervised](/content/unsupervised-cli/index.html): The Unsupervised CLI starts Claude Code with Finder, DeepWork, and analysis rules. Free to use, local by default, portable across agent workflows. (240 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/solutions/index.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (93 words) - [Finder for Teams — The Evidence Layer for Data Teams | Unsupervised](/content/finder-for-teams/index.html): Finder for Teams connects to your warehouse, searches for KPI-linked patterns, and ranks them by estimated financial impact, with evidence an analyst reviews before anyone acts. (260 words) - [Unsupervised - Agents You Can Trust to Run Unsupervised](/content/demo/index.html): Unsupervised builds the harnesses, workflows, observability, and checks that make agent autonomy trustworthy. (163 words) ## About Pages - [About | Unsupervised](/content/about/index.html): Unsupervised builds the systems that make agent autonomy trustworthy: harnesses, workflows, checks, and evidence. It started in enterprise data analytics. (196 words) ## Resources - [Full Page Index](/index.html): Browse all cached pages with rich metadata - [About This Cache](/about.html): Methodology, technical details, and usage guidelines - [XML Sitemap](/sitemap.xml): Machine-readable sitemap for crawler discovery - [Robots.txt](/robots.txt): Crawler directives - [Agentsite Network](https://agentsite.network/network.html): Public index of Agentsites and their machine-readable resources