- Specialization
- Not classified
- Business
Software Applicationsand 1 more business valuesLess
- Agent Tooling
- Market
- Software Development
- Ecosystem
- Agents
- Customers
Developersand 1 more customers valuesLess
- Enterprises
More details
- Organization type
- Commercial Company
- Ecosystem
- Agents
- Business
- Software Applications, Agent Tooling
- Specialization
- Not classified
- Market
- Software Development
- Customers
- Developers, Enterprises
- Delivery models
- Hosted Software, On Premise Software
- Revenue models
- Not classified
- Listing status
- Not classified
- Latest equity round
- Not classified
- Reported funding amount
- Not classified
About
HumanLayer builds an AI IDE, collaborative platform, and software-factory building blocks for engineering teams, serving customers from YC startups to Fortune 500 companies. Its differentiation is an opinionated, collaborative agent workflow designed to help teams ship complex software 2–3x faster while preserving code quality and architecture standards.
Market
HumanLayer competes in the AI-assisted software development and agentic coding platform market. It positions itself beyond autocomplete or a standalone coding agent as an integrated IDE, collaboration layer, and software-factory system for complex production codebases. Its differentiation is an opinionated Research-Plan-Implement workflow, multi-agent task collaboration, BYOK model access without a separate per-token bill, and human approval channels that emphasize maintainable code rather than AI-generated “slop”.
HumanLayer targets production software teams, from YC startups and small companies through Fortune 500 enterprises, particularly engineering organizations working in complex codebases. Its primary buyer personas are CTOs and VPs of Engineering seeking to increase AI-written software output while preserving code quality, architecture, and human oversight.
At a Glance
Problem
HumanLayer addresses the difficulty of using AI to accelerate work in complex codebases without sacrificing code quality, architectural discipline, or human oversight. Its stated promise is to help engineers ship two to three times faster while maintaining rigorous standards, targeting the particularly costly problem of solving large, ambiguous engineering tasks in million-line enterprise codebases. The main use case is coordinating a substantial change—such as an authentication refactor—across research, design, implementation, and review rather than merely generating isolated code snippets.
Product / Service
HumanLayer combines an AI coding IDE, collaborative cloud platform, and software-factory building blocks. Teams organize tasks, agent sessions, artifacts, and worktrees in shared workspaces; agents can produce and refine research documents, designs, plans, mockups, and code while humans inspect thinking messages, tool calls, and code changes. Local daemons can run multiple sessions in parallel, while cloud daemons continue long-running work after a laptop is closed, with activity synchronized through HumanLayer's API across web, desktop, and mobile interfaces.
The delivery model is subscription software with free, Pro, and custom Enterprise plans. Pro is listed at $100 per user per month and includes bring-your-own-key support for Claude Code, Codex, and other AI services, unlimited tasks and sessions, multi-repository workspaces, real-time human-agent collaboration, and remote daemons. This BYOK model avoids a separate HumanLayer per-token bill while giving teams a single environment for orchestration, collaboration, and visibility.
Market
HumanLayer competes across the AI coding IDE, developer-tools, collaborative coding, and software-factory orchestration categories. The available research does not identify a named direct competitor; instead, HumanLayer positions itself against the fragmented alternative of stitching together separate agent, task, artifact, and collaboration tools. Its differentiation is an opinionated, collaborative system for high-level orchestration that still exposes the underlying agent work and code changes.
The company is listed by Y Combinator as an active Fall 2024 Artificial Intelligence developer-tools company. HumanLayer says it is used by organizations ranging from YC startups to Fortune 500 companies and displays a set of customer logos, while its pricing page offers free, Pro, and Enterprise plans. It has disclosed a $500,000 pre-seed round led by Y Combinator in December 2024 and lists additional investors; however, the available evidence does not disclose revenue, so it should be characterized as commercially launched with early traction rather than assigned a confirmed revenue stage.
Founders & Leadership
Funding History
Y Combinator
Recent News
HumanLayer’s v0.178.0 release added draft pull-request creation from the GitHub tab or diff view and advanced metrics for paid plans. The release notes also describe clearer daemon-launch errors and npm tags aligned with release channels.
Y Combinator published a Launch YC announcement for HumanLayer on 2026-09-16, describing its platform for approval queues, escalations, and human oversight of automated processes. HumanLayer supports Python and TypeScript applications and lets AI agents request human feedback or approvals through Slack, SMS, email, and other channels.
A video about HumanLayer founder Dex Horthy’s software-factory experiments describes a July 2025 attempt in which an agent factory operated without anyone reading the code, ultimately falling apart.
HumanLayer explains harness engineering as the use of coding-agent configuration points to improve output quality and task-success rates. The post also describes integrations such as agents sending Slack messages, creating GitHub pull requests, and setting up preview environments.
HumanLayer project documentation listed a specialized Vercel AI SDK integration, Slack, email, CLI, and web interfaces for human interaction, plus TypeScript and Go SDKs.
Active Roles
1Business Model
HumanLayer appears to monetize access to its AI IDE and collaborative cloud platform through a bring-your-own-subscription model: customers connect Claude, Codex, or other AI subscriptions and API keys, with no separate per-token charge. Specific paid tiers and pricing are not disclosed in the available evidence.