Poolside is a private enterprise AI platform and frontier AI lab focused on software engineering. It builds foundation models, agents, and enterprise systems that let organizations run AI inside their own infrastructure, including virtual private cloud and on-premises environments, so code, prompts, and outputs remain under customer control.
The company combines model research with operational tooling for production software workflows. Poolside is designed for enterprises that need security, governance, auditability, and deployment flexibility while bringing AI into IDEs, browsers, terminals, APIs, and managed agent workflows. Its approach centers on long-horizon, tool-using software agents and the controls required to operate them in complex, regulated, and security-conscious environments.
Offerings, Capabilities, and Integrations
Poolside delivers a full-stack software engineering AI environment built around foundation models, agentic workflows, and enterprise controls. Its capabilities span code generation, explanation, testing, refactoring, modernization, incident investigation, and policy-aware development work grounded in private repositories, documentation, tickets, and internal standards. The platform supports single-agent and multi-agent workflows, sandboxed execution, policy controls, approval systems, auditability, and usage telemetry.
Poolside is built to fit into existing enterprise environments rather than require a rip-and-replace approach. It supports deployment on customer-controlled infrastructure, including on-premises and VPC-based environments, and it integrates with developer workflows through IDE surfaces, browser access, terminal workflows, APIs, and centrally managed agent configurations. Poolside also supports model connectivity through OpenAI-compatible endpoints and tested integrations with major cloud AI providers.
Products and Services
- Poolside Platform: Production-grade platform for running AI agents inside a customer’s security boundary, with deployment options across customer-controlled infrastructure and enterprise governance controls.
- Poolside Assistant: IDE-based agent experience for autonomous, multi-step software development tasks, with context gathering, tool use, code editing, approvals, and versioned change review.
- Poolside Chat: Browser-based interface for chat-driven software engineering interactions outside the IDE.
- Poolside Agent CLI: Terminal and automation interface for working with Poolside agents interactively, in scripts, and from ACP-compatible editors.
- Poolside Console: Administrative control plane for creating agents, auditing activity, managing models, tracking metrics, and administering organization settings.
- Developer API: OpenAI-compatible API for integrating Poolside models and workflows into internal tools, applications, and automation.
- Laguna M.1: Poolside’s most capable agentic coding model, designed for multi-step, tool-using software engineering workflows.
- Laguna XS.2: Open-weight agentic coding model optimized for lighter-weight, faster iteration and broad deployment flexibility.
- Malibu 2.2: Dense software engineering model that supports agentic and chat workflows for code understanding, debugging, editing, and validation.
- Point: Dense code completion model for inline and multi-line developer suggestions in supported editors.
- Forward Deployed Research Engineers: Embedded service model in which Poolside works with customer teams to assess use cases, build customized agents, deploy workflows, and operate them in production.
Target Customers
Poolside targets enterprises with meaningful software engineering complexity and strict control requirements. Its fit is strongest with product, engineering, and platform teams that need AI to work inside existing security boundaries, across private codebases, legacy systems, and heterogeneous infrastructure.
The company is particularly well aligned to organizations in highly regulated or security-sensitive environments, including financial services, defense, and industrial settings. It also addresses teams operating in multi-cloud, air-gapped, limited-connectivity, or compliance-driven environments where public SaaS coding assistants are difficult to adopt.
Cloud Integrations and Marketplace
- Amazon Bedrock: Poolside has an official Amazon Bedrock presence and supports AWS deployment paths that include Bedrock-based model access alongside customer-controlled AWS infrastructure options.
- Microsoft Azure Foundry: Poolside supports tested integration with Microsoft Azure Foundry as a chat model provider through its model management capabilities.
- Google Vertex: Poolside supports tested integration with Google Vertex as a chat model provider through its model management capabilities.
Key People
- Jason Warner: co-CEO & Co-Founder
- Eiso Kant: co-CEO & Co-Founder
- Margarida Garcia: COO
- Philip Drury: Chief Investment Officer
- Paul Almeida: Head of Public Sector
- Beatriz Lourenço: Head of People
Key Facts
- Headquarters: San Francisco, California, United States
- Employees: 150+
- Annual Revenue: Undisclosed
- Parent Company: None
- Subsidiaries: Fern Labs
- Publicly Listed: Private