Poolside
Also known as: Poolside, poolside.ai, poolside AI, Malibu, Point, poolside Assistant, Laguna
Enterprise coding platform pairing purpose built foundation models, IDE tools, and a governed multi agent orchestration console, deployed entirely inside customer infrastructure including air gapped and on premises environments.
Poolside is a full stack platform for enterprise software development, combining its own foundation models, developer tools, and a governed agent orchestration layer that runs entirely inside the customer's infrastructure. Founded in 2023 by Jason Warner, the former chief technology officer of GitHub, and Eiso Kant, the company trains its models with Reinforcement Learning from Code Execution Feedback, learning from actually running code rather than only reading it. Its flagship model Malibu handles code generation, refactoring, and test writing, while a smaller model, Point, drives low latency completion. Both can be fine tuned on a customer's own codebase, documentation, and knowledge bases.
On top of the models sits an agent platform. Through the Poolside Console, teams build and govern single and multi agent systems that run in isolated sandboxes, with an example agent taking a natural language specification, generating a working pipeline, pushing it to GitHub, building it through an API, and iterating until every step passes. Pipelines are code first and fully auditable, and meta pipelines review agent trajectories to refine how work gets done. The Console runs Poolside's own models alongside third party models such as Claude and GPT, configured per agent, giving teams real routing flexibility rather than a single fixed model.
Poolside's defining choice is sovereignty. Customers receive full model weights, not just API access, and deploy the entire stack into the environment they control, whether a cloud VPC, an on premises rack supporting hundreds of developers, a compact offline workstation for small classified teams, or their own NVIDIA hardware through Helm. Nothing leaves the boundary. The security posture targets demanding buyers: role based access control per agent, sandboxed execution with filesystem and network policies set before an agent starts, centralized secrets injected at runtime and redacted from logs, audit retention with SIEM integration, and government grade options including air gapped operation, a hardened operating system, and IL5 deployability.
The company is heavily backed. It raised a 500 million dollar Series B in 2024 led by Bain Capital, DST Global, and eBay at a 3 billion dollar valuation, and in 2025 Nvidia committed to invest up to 1 billion dollars, lifting the valuation toward 12 billion. Roughly 150 people work across the United States, United Kingdom, and France, with early defense customers including RTX Corporation. Poolside sells through Forward Deployed Research Engineers who embed with customer teams to ship agents into production. In early 2026 it released Laguna XS.2, a free open weight model for local agentic coding, alongside preview surfaces including a terminal agent.
Vendor details
Canonical URL
https://poolside.ai
Category
Coding agent
Subcategory
Enterprise coding foundation models and governed agent platform
Funding status
Independent and heavily funded. Founded in 2023 by Jason Warner, former chief technology officer of GitHub, and Eiso Kant, Poolside raised roughly 126 million dollars in seed and early rounds, then a 500 million dollar Series B in October 2024 led by Bain Capital, DST Global, and eBay at a 3 billion dollar valuation. In October 2025 Nvidia announced an investment of up to 1 billion dollars, reported to lift the valuation toward 12 billion. The company had about 150 employees as of late 2025, headquartered in San Francisco with a Paris presence.
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
Integrates natively into VS Code and Visual Studio for context aware completion and an in editor assistant, and ships a terminal agent for local use. Agents connect to tools through Model Context Protocol servers, with admin approved connections for GitHub, Linear, Slack, Jira, and Notion, and read knowledge bases backed by Git or Amazon S3. The platform exposes APIs and is available in Amazon Bedrock, and the Console manages repositories, sandboxes, credentials, and model providers. Agents act by editing files, running terminal commands, calling tools, and pushing to source control inside governed sandboxes.
In practice
You are in defense or government and cannot send code off premises. Poolside delivers full model weights into your air gapped environment, so every query and agent action runs inside your boundary with nothing leaving.
You want agents that do real work under control. Through the Console you build multi agent pipelines in sandboxes with per agent permissions, step limits, and full trajectory audit trails, so automation stays governed.
You want a model tuned to your own code. Poolside fine tunes Malibu and Point on your codebase, documentation, and knowledge bases, and can route across its own and third party models per agent.
Sources & related URLs
Related / legacy domains
Capability coverage
10.5 / 14 capabilities · 75%
| Integrations & Tool CallingIntegrates natively into VS Code and Visual Studio, connects agents to tools through Model Context Protocol servers with approved connections for GitHub, Linear, Slack, Jira, and Notion, reads knowledge bases backed by Git or Amazon S3, and acts by editing files, running commands, calling tools, and pushing to source control, broad named integrations with real action. | Full |
|---|---|
| Workflow OrchestrationLets teams build and govern single and multi agent systems in sandboxes through the Console, with an autonomous agent taking a specification, generating a pipeline, pushing it to GitHub, building it through an API, and iterating until every step passes, plus meta pipelines that refine workflows, genuine end to end multi agent orchestration. | Full |
| Knowledge Grounding & RAGTrains and fine tunes its Malibu and Point models on each customer's own codebase, documentation, and knowledge bases so the model reflects that team's practices, and gives agents read access to knowledge sources backed by Git or Amazon S3, deep codebase grounding that is a headline of the product. | Full |
| Human Oversight & GuardrailsEnforces runtime guardrails centrally: administrators set sandbox strictness, permitted Model Context Protocol servers, available tools, and hard step limits per run, with role based access control and secrets injected at runtime and redacted from output, a genuine runtime enforcement engine rather than advisory review. | Full |
| Security, Identity & GovernanceProvides a comprehensive governance matrix: role based access control per agent and team, sandboxed isolation, encrypted secrets, configurable audit retention with SIEM integration following NIST practices, full data residency with no egress, and government grade air gapped operation, a hardened operating system, and IL5 deployability. | Full |
| Observability & AuditabilityRecords a complete trajectory for every agent session, each tool call, file edit, reasoning step, and decision, searchable and exportable for compliance, alongside inference metrics including time to first token percentiles, agent dashboards, and SIEM propagation, comprehensive execution tracing, audit, and analytics. | Full |
| Memory & State PersistenceAdapts by fine tuning models on a customer's codebase and by letting agents read Git and Amazon S3 knowledge bases, both counted under knowledge grounding, and self improving meta pipelines are feedback loop adaptation, but no distinct persistent agent memory or checkpoint product is documented as first class. | Unable to verify |
| Deployment & Data ResidencyDelivers full model weights, not just API access, and installs the entire stack inside the customer boundary as a cloud VPC, an on premises rack, an offline workstation, or on customer NVIDIA hardware through Helm, fully air gapped if needed, a genuine self host and data residency capability across AWS, Azure, Google, and on premises. | Full |
| Prebuilt Agents, Templates & PacksShips platform primitives and a Console for building single and multi agent systems and code first pipelines, and Forward Deployed Research Engineers assemble agents for each customer, real custom agent scaffolding, though not a browsable library of prebuilt agents, templates, or packs that users adopt off the shelf. | Partial |
| Triggers & Channel CoverageReaches developers through the IDE, a terminal agent, a web assistant, the Console, and an API, and agents interact across the development ecosystem including source control and build pipelines, real trigger and channel coverage that centers on the software development workflow rather than broad external event channels. | Partial |
| Model Flexibility & RoutingRuns its own multiple purpose built models, Malibu, Point, and the open weight Laguna, and the Console also configures and routes third party models such as Claude, GPT, and open models per agent through reusable model provider settings, genuine model flexibility across both proprietary variants and multiple providers. | Full |
| APIs, SDKs & MCP ExtensibilityExposes public APIs, is available through a single API in Amazon Bedrock, hosts first class Model Context Protocol servers for agents, and supports reusable model provider connection settings, a strong API and MCP extensibility surface, though a documented public software development kit is not clearly surfaced. | Partial |
| Testing, Debugging & OptimizationTrains its models with Reinforcement Learning from Code Execution Feedback, so they learn from running code, generates and writes tests through Malibu, and runs autonomous agents that iterate until every step passes, a strong test gated and self correcting loop rather than a dedicated bug detection or quality engine sold on its own. | Partial |
| Browser & Computer UseRuns agents in isolated sandboxes that read and write the filesystem, execute terminal commands such as building, installing packages, and running git, and reach the network under allowlist policies, real code execution and computer use within governed containers, though not general autonomous browser automation as a core product. | Partial |
Pricing
Enterprise pricing is custom and not public, deployed in your environment through Forward Deployed Research Engineers. Poolside models are also available usage based in Amazon Bedrock. The Laguna XS.2 open weight model is free under Apache 2.0.
Enterprise deployment is a custom contract sized to the environment and developer count, delivered with Forward Deployed Research Engineers. Poolside models are also available usage based in Amazon Bedrock, priced per token. The Laguna XS.2 model is free to run locally under an Apache license.
Included quota
Enterprise: the full platform inside the customer boundary, including foundation models with full weights, the Console for agents and orchestration, IDE developer tools, sandboxed execution, governance and audit, and deployment across cloud VPC, on premises rack, offline tower, or customer hardware, with Forward Deployed Research Engineer support. Amazon Bedrock: managed Malibu and Point models billed per token. Laguna XS.2: free open weight model under Apache 2.0 for local use, with preview terminal and agent surfaces. Government options include air gapped operation, a hardened operating system, and IL5 deployability.
What is public
Public: the deployment models, the government grade options, availability in Amazon Bedrock, and the free open weight Laguna XS.2 release. Not public: enterprise list pricing, seat or usage minimums, and engagement fees.
Billing mechanics
Enterprise is a custom contract covering software, deployment, and Forward Deployed Research Engineer support, with infrastructure either provided or customer owned. Bedrock is pay as you go per token under AWS. The open weight model is free.
Cost watchouts
The real cost of an enterprise deployment includes infrastructure, whether a Poolside provisioned rack, a cloud VPC, or customer NVIDIA hardware, plus the Forward Deployed Research Engineer engagement. Bedrock usage bills per token and scales with developer activity. Air gapped and government deployments add hardware and compliance overhead.
Variable cost rationale
Exposure depends on the path. An enterprise deployment on committed or customer owned infrastructure behaves largely as a fixed cost, since inference runs on hardware the customer already pays for. Access through Amazon Bedrock is usage based and bills per token, so cost scales with developer activity. The blended picture is moderate: predictable for a committed on premises deployment, variable for Bedrock usage.
Additional watchouts
There is no public self serve price, so budgeting requires a sales conversation. Total cost includes infrastructure and services, not just software. Government and air gapped deployments carry additional hardware and compliance overhead. The free open weight model is not the same as the governed enterprise platform.
Overage / add-ons
Enterprise contracts are custom, so overage terms are negotiated. Amazon Bedrock access bills per token with no fixed seat, so cost scales directly with usage.
Sales call required
Yes — required for paid access
Free / trial
No public self serve trial of the enterprise platform; access starts with a sales conversation. A free path exists through the open weight Laguna XS.2 model, which runs locally under an Apache license, and through Amazon Bedrock usage based access.
Lowest paid plan
No public paid plan for the platform. The lowest cost paths are the free open weight Laguna XS.2 model run locally and usage based access to Poolside models in Amazon Bedrock, both short of the governed enterprise platform.
Commercial notes
Poolside is priced as a high touch enterprise platform, not a self serve tool, aimed at large organizations and the public sector that need models and agents inside their own boundary. Value concentrates where sovereignty, air gapped operation, and joint delivery matter. Teams wanting a low cost entry can start with the open weight model or Bedrock usage, then move to the governed platform.
Key ambiguities
No list price is public for the enterprise platform; cost depends on deployment shape, developer count, hardware, and the Forward Deployed Research Engineer engagement. Bedrock per token rates and any minimums are set through AWS. The boundary between free open weight use and paid platform capabilities is set by which components a team runs.
Cancellation / refund
Enterprise agreements are custom annual or multi year contracts negotiated directly. Bedrock usage is pay as you go under AWS terms. The open weight Laguna XS.2 model carries no contract.
Missing data
All enterprise list pricing, minimums, and Forward Deployed Research Engineer engagement costs are undisclosed. Bedrock per token rates are set through AWS and not restated here.
Related vendors
- Cognition — Maker of Devin, an autonomous AI software engineer
- Aider — Open-source, model-agnostic terminal coding agent that works…
- Anthropic Claude Code — Anthropic's terminal and IDE-native coding agent with sub-agent…
- Augment Code — AI coding assistant evolving into an AI-native engineering platform
- Blackbox AI — AI coding assistant for code generation and developer workflows
- Blink — Agentic AI app builder that turns a natural language prompt into a…