Pre-Seed Developer Tools Companies

Ashr offers a platform that evaluates AI agents by running tool calls, prompts, scoring outputs, and tracing execution paths.

Glue provides a software tool that allows developers and coding agents to build and modify user interfaces directly from code.

Sentrial offers an AI agent observability platform that monitors behavior, detects issues, suggests code fixes, and integrates via APIs.

Polymath builds AI training environments to develop and evaluate autonomous agents on long and complex software tasks.

Valgo develops software that quantifies risk for the insurance of physical AI systems.

Salus offers AI runtime controls that validate and block agent tool calls, enforce policies, and guide retries with feedback.

Sparkles offers a cloud environment where teams build, test, and preview code changes.

Nullstone is a developer platform that connects all engineering tools and allows developers to launch apps on cloud providers.

GenAI testing and validating startup

TensorPool makes ML model training effortless - just describe your job, and TensorPool handle GPU orchestration and execution.

Pig is a software development company that develops API for automating Windows desktops with AI.

Evolvable provides software that automatically writes, tests, fixes, and improves codebases continuously.

Interplay connects and automates design and code workflows to align your entire product team around a single source of truth

Compresr is a platform that compresses text for LLMs, saving tokens and costs, and includes an API, a Python SDK & adjustable compression.

Replicas is a software that provides virtual machine workspaces for AI coding agents to run and manage software development environments.

Confluence Labs develops AI systems for data-efficient reasoning and program synthesis to support research in science and engineering.

Watermelon offers an open-source copilot for code review

MARC is an AI startup dedicated to real estate asset management.

Mohi provides a platform that analyzes and clarifies complex AI agent workflows to help optimize and troubleshoot autonomous processes.

Guepard: Serverless databases with Git Superpowers for Teams and Agents

Cyberdesk provides virtual desktops for deploying AI agents using simple code.

Cactus enables AI model execution on mobile platforms through an open source inference engine with support for multiple app frameworks.

BootLoop enables AI agents to write, test, and debug firmware directly on the intended hardware.

Keystone is an AI-native testing platform that helps developers build, maintain, and run end-to-end tests at scale.

Async offers a coding platform that enables the development of end-to-end apps, encompassing building, testing, and deployment.