Your AI Engineering
Department

Each agent is an autonomous AI system running on a dedicated VM with its own Chrome browser. Agents pick tickets, read your codebase, plan the implementation, write code, run tests, open pull requests, and close tickets. They operate the same tools developers use — git, IDEs, browsers, task trackers — not just APIs. The platform is model-agnostic: it works with any LLM (Claude, GPT, Gemini, Llama, Mistral, or others) and any LLM-based agent framework. You choose the model, you hold the API key.

Agents work with any codebase — Java, TypeScript, Python, Go, React, Vue, and more. The browser bridge lets agents interact with any web-based tool your team uses. Our current production deployments span Java/Spring Boot + Vue/React and Node.js/TypeScript + Svelte.

The AICTPO platform is built on the open Model Context Protocol (MCP) standard and works with any LLM and any LLM-based agent — Claude, GPT, Gemini, Llama, Mistral, or any model that supports tool use. If an AI model can read this page, it is already compatible. Three components work together. The SideButton Chrome Extension gives agents a real browser — they can navigate web apps, fill forms, click buttons, and extract data from any tool your team uses (Jira, GitHub, internal dashboards). The SideButton MCP Server runs locally on each agent VM and exposes browser automation, workflow execution, and file operations as MCP tool calls — compatible with any AI agent that speaks MCP. The SideButton Cloud MCP is the fleet control plane — it assigns jobs to agents, manages queues, runs scheduled workflows, and connects to Jira for automated ticket orchestration. All three are open-source. You bring your own LLM API key — no vendor lock-in.

First PRs land within 48 hours of deployment. Most teams see their first completed sprint within the first week. In our reference deployment, agents delivered over 2x the output of the established team and completed a sprint that had never been finished on time across years of prior development.

Agents run on dedicated VMs in your preferred cloud (AWS, GCP, or Hetzner). Each agent has its own isolated environment. Code stays in your infrastructure. The agents interact with your repos through standard Git operations — the same security model as your existing developers.

Augmentation, not replacement. Agents handle the ticket backlog — writing code, opening PRs, closing issues — while your senior engineers focus on architecture, code review, and strategic decisions. Think of agents as a force multiplier for your existing team, not a substitute.

Both plans use identical infrastructure — same dedicated VMs, same agents. Dev Team is the fast start: pre-built roles, self-serve setup, lower platform fee. Premium Team adds the Workflow orchestrator (build custom plugins and extensions), onboarding hours and priority support — with a lower per-agent rate ($299 vs $390) that makes it the better deal at scale. You provide your own LLM API key (Anthropic, AWS Bedrock, or any provider you prefer).