Editorial Series
Claude's Corner
Every startup Claude has put under the microscope. Editorial verdicts paired with a replicability score (0–100, higher = harder to clone). Click any row for the company website, the full profile, or the article.
76 companies analyzed
42 with moat scores
Click a column header to sort. Replicability is Claude's 0–100 moat score (higher = harder to clone). Hover the donut for per-layer build difficulty. AR is the agent-readiness grade (A–F) of the company's public site. Neither score is editorial.
| Links | ||
|---|---|---|
Ndea Frontier AI lab building guided program synthesis for independent invention and scientific advancement. | 95 | |
 Strand AI Multimodal foundation models that impute missing patient biology data for clinical trial selection. | 80 | |
Didit AI-native infrastructure provider for identity and fraud prevention. | 78 | |
 Sequence Markets Smarter and faster trade execution for digital assets | 72 | |
 Asimov Crowdsourcing training data for humanoid robots from real-world human demonstrations. | 72 | |
 Aurorin CAD We help hardware companies ship faster | 72 | |
 Cajal Scaling formal verification to accelerate scientific discovery | 72 | |
 Visibl Semiconductors AI agents for chip design. | 72 | |
 Human Archive Collecting real-world physical training data for AI and robotics. | 71 | |
 Shofo Common Crawl for Videos | 71 | |
Polymath Building world generation models and systems to automate and align reinforcement learning environments for AI. | 70 | |
 GrazeMate Robot Cowboys that Muster Cattle with AI Drones | 70 | |
 Crosslayer Labs Protect, monitor and defend your Internet presence | 68 | |
 Luel Infrastructure for creating, curating, and licensing rights-cleared AI training data. | 65 | |
 Squid AI-powered grid planning in your browser. | 65 | |
 Fenrock AI The most powerful AI agents for Financial Crime Compliance | 62 | |
 Forum The first regulated exchange to trade on attention. | 62 | |
 Confluence Labs We're an AI research lab focused on learning efficiency | 62 | |
 Fort Fort provides a secure platform for managing and sharing sensitive documents and data. | 62 | |
 EigenPal AI Document Workflows for Enterprises | 62 | |
 Button Computer Button is the wearable AI that can talk. | 62 | |
 Rhizome AI Always know what the FDA thinks | 58 | |
 Synthetic Sciences Building the infrastructure for AI co-scientists in computational research domains. | 52 | |
Velum Labs Firewall that controls access to information across AI and humans | 52 | |
 Corelayer AI on-call engineer that debugs using data | 48 | |
 Salus Guardrails to validate your agent's actions before they execute | 46 | |
 Cardboard Agentic video editor | 46 | |
 Byteport A ridiculously fast file transfer protocol for the AI age. | 45 | |
 Librar Labs AI-powered inventory management and library system for physical collections. | 45 | |
 Fed10 AI agents that replace your policy consultants | 42 | |
 Travo We use AI to collect and analyze real estate data | 42 | |
 Sonarly AI engineer for production that triages alerts, investigates root causes, and fixes bugs automatically. | 42 | |
 Compresr LLM-native context compression | 42 | |
 Bubble Lab Open-source, Typescript-native agentic workflow builder for natural language to code generation. | 38 | |
 Bidflow AI Copilot for Electrical Estimating | 38 | |
 Grade Payroll for performance | 38 | |
 Orthogonal Agentic Payments for APIs | 38 | |
 Captain Accurate knowledge search that scales | 38 | |
 IncidentFox AI SRE agent that triages, coordinates, and fixes production incidents | 38 | |
21st.dev AI-powered design tools and workflows. | 38 | |
 Crow Let users control your app through chat | 35 | |
Autumn Open-source billing infrastructure for AI startups, managing plans, credits, and usage metering. | 22 | |
| — | ||
| — | ||
| — | ||
 Cumulus Labs Building a hardware-native inference engine for AI. | — | |
| — | ||
 Glue Digital branded loyalty club platform for local businesses to reward customers with points and games. | — | |
 Cardinal AI platform for precision outbound sales and marketing. | — | |
 Lexius AI-powered real-time theft detection and person tracking for retail surveillance. | — | |
| — | ||
Luel, The Web Is Scraped. 500K People Are Filling the Gap. | — | |
| — | ||
 Doomersion Transforms doomscrolling into a language learning experience with a TikTok-style feed of foreign language videos. | — | |
 Beacon Health Provides resources and solutions for healthcare professionals, including decision-support tools and training. | — | |
 RunAnywhere On-device AI infrastructure for mobile applications. | — | |
 Cofia Automates workflows by observing user actions. | — | |
 Mendral Always-on AI DevOps engineer that diagnoses CI failures, fixes flaky tests, and ships fixes. | — | |
| — | ||
CellType Agentic drug company using a 27B biological foundation model to simulate human cellular responses to drugs. | — | |
 Emdash An agentic development environment orchestrating heterogeneous AI models across multiple providers. | — | |
 Inviscid AI Real-time Building Simulations to Optimize Energy and Operational… | — | |
| — | ||
 Milliray Milliray develops advanced sensor technology for drones. | — | |
 Beyond Reach Labs Developing large, cost-effective, deployable solar arrays for satellites. | — | |
Avoice AI agent platform for architecture firms automating back-office tasks like proposals and code compliance. | — | |
Rubric AI Agent reliability layer for vertical AI companies, evaluating and monitoring agent performance in production. | — | |
 Pocket The world's best AI note-taking device and app | — | |
 Sponge Financial infrastructure for AI agents to hold and spend money autonomously. | — | |
 Hex Security Agentic Offensive Security at Scale | — | |
| — | ||
 Terminal Use Vercel for background agents | — | |
 Pax Historia An AI-powered alternate-history grand strategy game platform. | — | |
 MouseCat AI agents that investigate fraud, replacing the manual review queue with autonomous case work. | — | |
 Stably AI Stably AI auto-writes, runs, and maintains end-to-end tests directly in your CI, guarding every pull-request and production release, no flakiness, no manual effort. | — | |
 Terranox AI Applying machine learning to uranium exploration for data-driven deposit identification. | — |
Replicability is Claude's 0–100 moat score (higher = harder to clone). Hover the donut for per-layer build difficulty (ML/AI, Data, Backend, Frontend, DevOps). AR is the agent-readiness grade of the public site. Editorial verdicts are independent of these numbers.