.webp)
Startup Intelligence Database
Discover what the world's best startups are building before it's announced publicly.
RAG
Data Extraction
Computer Vision
Speech AI
Model Evaluation
Kinect
Personalized AI storefronts that adapt PDPs for each shopper.
Brands need owned storefront intelligence before AI agents become a buying channel and third-party assistants tax product discovery from both sides.
Computer Vision
Model Evaluation
Workflow Agent
Natural Language Processing
Indexable
Forkable VM sandboxes for AI agents.
As AI agents start doing real work, stateful VM branching becomes core infrastructure, and Indexable is early, opinionated, and built by a systems engineer who knows what that layer has to do.
RAG
Data Extraction
Speech AI
Model Evaluation
Multimodal AI
AgentPhone
Gives AI agents real phone numbers for calls and texts.
AgentPhone is selling the missing telecom layer for agents at the moment every agent platform needs real-world phone I/O.
RAG
Data Extraction
Computer Vision
Model Evaluation
Personalization
Arden
Automates SOX control testing with audit-grade AI agents.
AI-native SOX testing is a fast-funded category, and Arden's wedge is the workpaper itself rather than another dashboard sitting next to AuditBoard.
Model Evaluation
Prediction
Personalization
Fraud and Risk Scoring
Anomaly Detection
MochaTrade
US stock perps for non-U.S. retail traders.
India has enormous F&O appetite and limited US market access. MochaTrade pairs local INR rails with self-custody to package 24/7 stock-linked perps for traders the incumbents cannot reach.
RAG
Model Evaluation
Workflow Agent
Natural Language Processing
Manicule
AI agents and human editors build docs for developer-tool startups
A tiny YC team is turning docs maintenance into sticky workflow software through a services wedge, agentic QA tooling, and devtool logos that buyers recognize.
RAG
Data Extraction
Model Evaluation
Simulation
Fraud and Risk Scoring
Astraea
Automates clinical-trial biometrics into FDA-ready outputs.
If Astraea can make FDA submission math boring, small biotechs get CRO-grade biometrics without the CRO drag.
Data Extraction
Model Evaluation
Personalization
Recommendation Systems
Workflow Agent
FlowScope
Map business processes, then automate them with AI agents.
Task mining collapsed into one motion with build and maintenance lets FlowScope sell outcome-based automation before buyers commit to hiring an AI team.
RAG
Data Extraction
Model Evaluation
Fraud and Risk Scoring
Workflow Agent
Gigacatalyst
Embedded AI app builder for SaaS customization
SaaS customization is shifting from services tickets to user-built microapps, and Gigacatalyst is positioning to be the embedded workflow layer that vertical software vendors offer their own customers.
Model Evaluation
Prediction
Fraud and Risk Scoring
Anomaly Detection
Synthetic Data
Klaimee
Liability coverage and certification for autonomous AI agents
Agentic AI needs a new risk backstop, and Klaimee pairs YC timing with founder insurance ops experience to build the certification and policy layer.
Data Extraction
Speech AI
Prediction
Workflow Agent
Natural Language Processing
KugelAudio
On-prem multilingual TTS for enterprise voice agents.
Europe needs a sovereign TTS layer. KugelAudio has real model work plus an ElevenLabs-compatible API that gives buyers a clean migration path.
RAG
Computer Vision
Model Evaluation
Simulation
Multimodal AI
TesterArmy
Test web and mobile apps with an AI QA agent before users find bugs.
AI-coded apps need AI-native QA, and TesterArmy attacks the review bottleneck directly inside the PR loop where the bug actually appears.
Model Evaluation
Optimization
Personalization
Fraud and Risk Scoring
Workflow Agent
Amboras
AI-native e-commerce platform that builds and runs online stores.
If Shopify made merchants stitch apps together, Amboras is betting the next store OS is an agent that builds, edits, and optimizes the shop on the merchant's behalf.
RAG
Data Extraction
Model Evaluation
Workflow Agent
Natural Language Processing
Modern
Resolves IT requests with agentic workflows across enterprise systems.
YC timing, founder fit, and a ServiceNow-sized pain point make Modern a sharp early bet if agents can execute IT work safely inside enterprise systems.
Data Extraction
Computer Vision
Model Evaluation
Prediction
Optimization
InsForge
Backend platform for agent-built apps
If agents become the default app builder, the backend layer needs an API they can operate end-to-end, and InsForge is early to that control-plane wedge.
RAG
Model Evaluation
Optimization
Workflow Agent
Natural Language Processing
ReasonBlocks
Agent runtime that cuts repeated failures and token waste.
Agent spend is moving from chat to long tool loops, and ReasonBlocks attacks the ugly part of the bill: repeated failures that current frameworks only notice after the damage.
RAG
Data Extraction
Computer Vision
Model Evaluation
Prediction
Framewise Health
Turns discharge data into patient-specific recovery videos.
Owning the discharge moment turns patient education from a nurse labor sink into measurable readmission ROI that hospitals can price against CMS HRRP penalties.
Data Extraction
Computer Vision
Model Evaluation
Fraud and Risk Scoring
Multimodal AI
Arctic Health
Automates provider credentialing and payer contracting.
Credentialing is expensive, slow, and rules-heavy. Arctic sells outcome ownership into a market that has been ready to swap spreadsheets for reviewed automation for years.
Data Extraction
Model Evaluation
Fraud and Risk Scoring
Workflow Agent
Natural Language Processing
Clawvisor
Authorizes AI agent actions without exposing user credentials.
Agent permissions are turning into a real security bottleneck, and Clawvisor combines strong founder-market fit with open-source velocity to attack that layer early.
Computer Vision
Model Evaluation
Multimodal AI
Natural Language Processing
Intelligence Factory
General-purpose manipulation models for robots
If tactile human demonstrations become scarce training fuel, Intelligence Factory is positioned to own the data layer that makes robot hands actually useful.