Helpin

You start either by signing up for Helpin Cloud, which opens with a 14-day trial of the Growth plan and no card required, or by installing the self-hosted edition with the Helpin command-line tool. The CLI installs the stack (Docker Compose, PostgreSQL, file storage and an agent runtime), checks its health with a doctor command, reads service logs, and handles backups, restores and upgrades. You can evaluate it locally and later move it to a public server, where the CLI generates the Caddy configuration.

Next you bring your work in. Chat and email land in shared team inboxes, help articles go into knowledge spaces and a public help center, and tasks, stories, epics and sprints live in projects. Meeting transcripts are summarised into notes and action items, and customer records and deals sit in the CRM.

From there you hand work to agents. You can ask Helpin AI a question about a customer or a delayed project, and it searches conversations, meeting notes, docs and tasks, pulling in specialist agents when needed. A bug report can flow from Echo in support to Scribe for task planning, Forge for the code change and pull request in GitHub or GitLab, and Lens for review, while Quill updates the guide. Each agent only uses the tools you grant, including external tools connected through MCP (in beta), and approval rules decide what it may do without a human sign-off.

• •Shared inbox with a support agent - Chat and email support run in shared team inboxes with saved views and assignment. The Echo agent answers from your latest docs, past conversations and connected tools such as billing, error logs or issue trackers, checks back in after a reply, and hands the conversation to a person when it should. Growth adds AI conversation routing and round-robin assignment.
• •Specialist agents and custom agents - Eight built-in agents cover support, planning, task planning, coding, code review, docs, CRM and marketing. On Growth and in the self-hosted edition you can describe new agents and automation flows in natural language, review their instructions, tools and steps, and schedule agent runs or trigger them from task changes and GitHub or GitLab events.
• •Coding agents tied to customer requests - Forge writes the change, runs the tests and opens a pull request in your connected repository with the original request attached, and Lens checks the proposed code for bugs, missing tests and unmet requirements before your team decides. Projects support tasks, stories, epics, sprints, roadmaps and objectives.
• •Knowledge base and help center - Product guides and internal docs spaces feed the agents, and a public help center can run on your own domain. Quill prepares guide updates from customer questions and product changes, using browser screenshots and recordings where steps need showing, and your team reviews and publishes. Growth adds a multilingual help center and AI article translation.
• •Meetings and CRM in the same history - Meeting transcripts become summaries and action items that agents can turn into plans and follow-ups. The CRM holds contacts, deals, signals and playbooks, and Beacon finds accounts that need attention, updates records and drafts follow-ups using the account conversations and open work.
• •Self-hosting and developer access - The AGPL-3.0 community edition runs on your own infrastructure with no telemetry by default, unlimited users and your own provider keys (OpenAI, Anthropic, OpenRouter or compatible endpoints). Developers get a widget SDK, a Helpin MCP server for connecting AI clients, and the operations CLI.

Helpin fits small and mid-sized software companies where the same few people handle support, product and sales, and where the gap between a customer complaint and a shipped fix is mostly lost context. A SaaS founder can let Echo take routine questions, have Forge draft fixes for reported bugs, and keep the help center current through Quill without hiring a separate support operations team.

Engineering-led teams that already live in GitHub or GitLab will get the most out of the coding and review agents, since every pull request can carry the customer request that caused it. Support leads who are tired of per-seat and per-resolution pricing in tools like Intercom will notice that Cloud plans charge per workspace with unlimited teammates and an included AI allowance.

Teams with data residency or compliance requirements, and developers who simply prefer to run their own stack, can self-host the open-source edition and pay only their own servers and AI provider. Be aware the self-hosted community edition is still marked beta, and commercial licensing outside AGPL, deployment help and support terms are handled through an Enterprise agreement. Customer success and sales staff at smaller B2B companies can also use the meeting notes and CRM agent to keep commitments from slipping between calls.