Skip to content
fromHello
Best · Open source

Best open-source customer engagement platforms (2026)

The closest thing to a complete open-source customer engagement platform in 2026 is Dittofeed, now in maintenance mode; Mautic covers the most ground for marketers; Chatwoot and Novu each cover half the job — conversations and notifications. Laudspeaker is no longer maintained. fromHello pairs an AGPL platform with an AI growth team, in early access.

Updated 6 Sept 20269 min readBy fromHello
Join the waitlistStar on GitHub
Key takeaways
  • No single open-source project covers the whole customer engagement job — profiles and events, segments, journeys, and email, SMS, push, and in-app messaging. Every open-source pick is a scope decision first.
  • The category is shrinking, not growing: Laudspeaker declared itself unmaintained in July 2026, LimeJourney never cut a release, Castled.io's code is gone entirely, and Dittofeed's maintainer confirmed in August 2026 that feature development is halted.
  • The two healthiest projects here run adjacent lanes: Chatwoot covers engagement as conversations, Novu covers notification delivery — neither runs lifecycle journeys on behavioral segments.
  • fromHello pairs the AGPL platform with eight AI specialists who run it, but it is in early access — weigh it as a waitlist today, not a download.

Search for an open-source customer engagement platform and most of what you find is dead, adjacent, or both. Two of the category's names have gone dark — Laudspeaker's final commit is its own abandonment notice, and Castled.io's repositories are deleted entirely — while the maintainer of Dittofeed, the most complete survivor, confirmed in August 2026 that feature development is halted. What remains is worth mapping honestly: one frozen-but-working CEP, one veteran automation suite, two healthy projects that each cover half the job, and our own early-access entry — plus Laudspeaker itself, mapped so you don't have to rediscover why it's off the table. This page sorts them by how much of the customer engagement job — profiles and events, segments, journeys, multi-channel messaging — each actually covers, with licenses and watch-outs stated plainly.

How we chose

The bar for this list: a genuinely open-source license on the platform itself (MIT SDKs for a closed API do not count), a project alive enough to run in production, and real coverage of at least part of the customer engagement job. Laudspeaker stays on the list despite failing that bar — readers search for it, and its notice is part of the category's story. We verified every claim against primary sources during the first week of September 2026 — each project's own repository, license file, docs, and pricing pages — and linked them below. Adjacent lanes are named as adjacent rather than stretched: notification infrastructure is not lifecycle messaging, and a support inbox is not a journey engine. We also checked the names readers ask about — erxes, a business-OS suite where engagement is one plugin among many; Tracardi, licensed MIT plus Commons Clause — source-available rather than open source; Tiledesk and Notifo, smaller neighbors of Chatwoot and Novu — and held the list to projects attempting the CEP job itself. We build fromHello, so it is listed last and held to the same standard, including the parts that are not ready yet.

The shortlist
  1. 01
    DittofeedMIT

    Best for: Teams that want the most complete open-source CEP data model — segments, journeys, email, SMS — and accept a project in maintenance mode.

    Strengths
    • The most CEP-shaped open-source project to date: dynamic segments, a visual journey builder, broadcasts, and a Segment-compatible API for profiles and events.
    • Real channel breadth for a self-hosted stack — email through SendGrid, SES, Resend, Postmark, or SMTP, and SMS through Twilio or SignalWire, plus webhooks for everything else.
    • Still runnable and still sold: the MIT core self-hosts with Docker Compose, and the hosted cloud remains open with plans from $75/mo.
    Watch-outs
    • Officially in maintenance mode: in August 2026 the maintainer wrote that feature development has been halted, with infrequent opportunistic bug fixes. The last stable release dates to December 2025.
    • Mobile push and in-app messaging were never finished — the docs list both as future work, and with feature development halted they are unlikely to arrive in the open-source edition.
    Compare in depthVisit site
  2. 02
    MauticGPLv3

    Best for: Marketing teams that want the most ground covered in one actively maintained open-source suite and can run a cron-driven PHP stack.

    Strengths
    • A decade-plus project still shipping: version 7.2 arrived in September 2026, with dynamic segments, a visual campaign builder, landing pages, forms, and A/B tests.
    • The widest channel spread in open source — email in core, SMS via a Twilio transport, web and mobile push through your own OneSignal account, and on-site messages via Focus Items.
    • Genuinely free under GPLv3 with no paid core tier, and community-governed — its AI stance is deliberate: bring your own models via plugins rather than a hosted service.
    Watch-outs
    • It is marketing automation grown toward engagement, not an event-native CEP: behavioral tracking centers on page visits, forms, and email activity rather than an arbitrary product-event stream, and segments rebuild on cron rather than in real time.
    • The stack is yours to operate — PHP, MySQL, cron jobs, upgrades — and push and SMS need third-party accounts before they deliver anything.
    Compare in depthVisit site
  3. 03
    ChatwootMIT

    Best for: Teams whose customer engagement means conversations — a shared inbox for chat, email, and WhatsApp — rather than lifecycle campaigns.

    Strengths
    • One of the two healthiest projects on this page — roughly 36,000 GitHub stars, releases every few weeks, and an active company behind it.
    • The strongest open-source support surface: live chat, email, WhatsApp, Instagram, Telegram, and more in one shared inbox, self-hosted on Rails and Postgres.
    • Captain, its shipped AI layer, drafts replies, automates first responses, and answers FAQs — available on paid plans, including self-hosted ones.
    Watch-outs
    • It covers the conversations half of engagement, not the lifecycle half: campaigns are single messages, automation rules have no wait or drip steps, and there is no journey builder or product-event stream.
    • The open-core line is real: Captain, SSO, and SLA features sit outside the MIT core, on paid per-agent plans.
    Visit site
  4. 04
    NovuMIT

    Best for: Product and engineering teams that need notification delivery — in-app inbox, push, email, SMS — wired into their own app, not a marketing platform.

    Strengths
    • Purpose-built notification infrastructure, actively developed at scale: roughly 40,000 GitHub stars, an embeddable in-app Inbox component, and dozens of delivery providers.
    • Multi-step workflows with delay, digest, and throttle steps cover the notification job properly, triggered from your code with per-step conditions.
    • Novu Connect, shipped in June 2026, wires AI agents into Slack, Teams, WhatsApp, Telegram, and email with human approval built in.
    Watch-outs
    • There is no marketing layer: no dynamic segments recomputed on events, no lifecycle stages, no campaign suite — targeting is subscribers, topics, and step conditions.
    • Community self-hosting of the MIT core is deliberately thinner than cloud: one team member, no SSO or RBAC, no email activity tracking, and a MongoDB-based stack outside the usual Postgres toolchain.
    Compare in depthVisit site
  5. 05
    LaudspeakerAGPLv3

    Best for: No one starting today. Laudspeaker was the closest open-source Customer.io alternative — and its README now says it is no longer maintained.

    Strengths
    • The right idea, executed openly: a visual journey builder, segment builder, and event-triggered messaging across push, email, SMS, and webhooks under an AGPL core.
    • Its mobile SDKs and hundred-million-user ambitions made it the most credible open-source Braze challenger of its generation.
    • The code is still public, so an existing install keeps running and the repository remains a reference for how an open-source CEP fits together.
    Watch-outs
    • The project states plainly that it is no longer under active development or being supported — that notice is the repository's final commit, from July 2026.
    • The website is unreachable and the hosted cloud's infrastructure has been removed, so there is nothing to fall back on and nowhere to migrate up to.
    • The last release dates to June 2024; in-app messaging was still marked in-progress on the roadmap when development stopped.
    Visit site
  6. 06
    fromHelloAGPLEarly access · waitlist

    Best for: Small teams that want the open-source platform and an AI growth team to run it — once early access opens.

    Strengths
    • The only option here that pairs the open-source platform with eight AI specialists who set strategy, build journeys, write copy, and run experiments — you approve, edit, or auto-ship.
    • A complete platform underneath rather than a chat layer over one: email, SMS, in-app, web push, ad audiences, a 16-node visual journey builder, real-time segments, GDPR consent tracking, and a first-party JS/TS SDK.
    • AGPL-licensed and self-hostable on Postgres + Redis, so your customer data stays on your own infrastructure.
    Watch-outs
    • The hosted product and the agents are in early access behind a waitlist — you can adopt Dittofeed, Mautic, Chatwoot, or Novu today.
    • Hosted pricing is announced at general availability; self-hosting the open-source platform is free.
    Join the waitlist
FAQ

Common questions

  • What is an open-source customer engagement platform?

    A customer engagement platform holds customer profiles and events, computes segments, and runs journeys across email, SMS, push, and in-app from one place — open source means you can read the code and run it on your own infrastructure. The test that separates it from its neighbors: if it cannot send the message, it is a CDP, not a CEP; if it cannot target on behavior, it is notification infrastructure.

  • Why are there so few open-source customer engagement platforms?

    Because a CEP is really four products in one — a data store, a segmentation engine, a journey runtime, and channel delivery — and the delivery half depends on paid third-party services whether the code is open or not. The recent record is blunt: Laudspeaker declared itself unmaintained in 2026, LimeJourney never cut a release, Castled.io deleted its code, and Dittofeed's maintainer confirmed in August 2026 that feature development is halted. The projects that thrive pick one half of the job.

  • What happened to Laudspeaker?

    Its README now opens by stating the project is no longer under active development or being supported — a notice added in July 2026 as the repository's final commit. The website is down, the hosted cloud's infrastructure has been removed, and no successor or recommended fork exists. Existing self-hosted installs keep running, but nobody is patching.

  • Does PostHog replace a customer engagement platform?

    Not yet. PostHog's Workflows entered beta in November 2025 with email; SMS through Twilio and push followed — push still in beta. It is a real step toward messaging, attached to its analytics suite — but it is a cloud beta with no in-app journeys, so re-evaluate it at general availability rather than treating it as a self-hosted CEP today.

  • Can I assemble a customer engagement platform from smaller open-source tools?

    Yes, and for narrow needs it is the pragmatic answer: Listmonk, Keila, or Plunk for email, Chatwoot for conversations, Novu for notifications, PostHog for analytics. The cost is the glue — identity, consent, and suppression logic live in none of them, so every tool you add multiplies the integration you own. Past a couple of channels, the glue usually costs more than the platform.

  • Should I still pick Dittofeed in 2026?

    With open eyes. It remains the most complete open-source CEP — still runnable and still sold — but the maintainer said in August 2026 that feature development has halted, mobile push and in-app never shipped, and the last stable release is from December 2025. Fine for a contained deployment you can maintain yourself; risky as the center of a growing stack.

Sources

See the platform the team runs.

Join the waitlistStar on GitHub
Related comparisons
Related guides
Best-of & alternatives
Switching tools
How to switch email platforms without losing deliverability
Early access

Put your growth teamon autopilot.

Early access opens gradually, so the team tunes to real use cases. Small teams with big ambitions go first.

Not ready to share an email? It's open source. Run it yourself today. View on GitHub

No spam. One email when your spot opens. Unsubscribe at any time.