Show HN: Libredesk – Open-source customer support desk. Single binary app

[avr5500]

Libredesk is a 100% free and open-source customer support desk, the backend is written in Go and the frontend is in Vue JS with ShadnCN for UI components.

Unlike many "open-core" alternatives that lock essential features behind enterprise plans, Libredesk is fully open-source and plans to always stay this way.

It's in alpha (v0.1.0) right now, but there’s a working demo available. I built this because I wanted a truly open and self-hosted alternative to platforms like Chatwoot, Intercom, and Zendesk.

Would love feedback, suggestions, and thoughts from the community.

GitHub: https://github.com/abhinavxd/libredesk

Demo: https://demo.libredesk.io/

[firefoxd]

Kudos! We've encountered several customers that hated their help desk but couldn't move away. We've even considered building our own since we were in the business. But here comes the hard and most important part that Zendesk and the like do that makes all the difference: Integration.

You need support integration with all the popular tools people use their help desk with. Shopify, ebay, whatsapp, woocommerce and all the seemingly unrelated software people use to support their e-commerce business. That's where we gave up, but I hope you find a solution for it.

[aarreedd]

Surprised no one has mentioned FreeScout. It's a solid open-source helpdesk: https://freescout.net/.

They have cheap, one-time purchase plugins. Happy to pay a bit to keep the project sustainable.

I'd love to see an open-source Trello alternative that as well. There are a few out there, but nothing that seems actively maintained.

[alchemist1e9]

https://github.com/RotherOSS/otobo

Otobo might be a good other option. It’s a refreshed and expanded version of the classic OTRS. Yes perl, but in this space has an extensive track record.

[mattsimpson]

Damn impressive start. Kudos! I am sure that if you posted a to-do list and roadmap to 1.0 you would get traction from the community to help finish it off, if you want. Hell, at a solid 1.0 I might be able to justify a switch from a popular commercial app and use that budget to fund a bit of development towards the project.

[avr5500]

Thank you! Adding a todo and a roadmap to v1.

[robertlagrant]

Potential roadmap item: bidirectional sync with Jira (and others) so support requests can create/be linked to Jira (etc) tickets, and see updates from them.

[phrotoma]

While we're thinking about it, something I always wanted when I was working support. The abilitiy to "upvote" a bug from within a support case. That way a PM can sort open issues by number of users who are griping about it.

[robertlagrant]

Yeah! Or at least have a count of the number of customer requests linked to the same ticket.

[rsyring]

We looked years ago for simple helpdesk that would link to GitHub issues so that CS was working in the helpdesk tool but could easily create issues when a dev is involved.

[teruakohatu]

When I last looked, I couldn't find any truely open source solutions that could practically replace even a shared inbox. This looks like it could. Well done to the developer.

And the paid options had high costs even on the most basic plans.

[potamic]

Who are your target users? Why not use an embedded db like sqlite and make it truly "single binary"?

[codazoda]

I would dig this

[grzaks]

Excellent! I was looking for a lightweight OSS alternative to Freshdesk/Zendesk and, honestly, didn’t find any worth deploying. Yours looks very promising, and we’ll definitely assess it.

[yorwba]

I don't know how lightweight you're looking for, but you might also want to have a look at Zammad: https://github.com/zammad/zammad I worked with a self-hosted deployment of it before, and it was reasonably low-maintenance.

[pentacent_hq]

Great job, that's looking very nice and polished already! Are you also planning to introduce a SaaS version of it?

[avr5500]

No plans, cloud deployment platforms like Railway can take of it

[nh2]

How do you plan to fund the development, if at all?

For a project like this, a hosted version might be a nice idea if you eventually want to do something else than putting your free time into support and maintenance. And many users will appreciate that the hosted version takes care of DB migration and backups for them.

A hosted version need not necessarily target maximum money making. You could run it as a nonprofit whose goal is to ensure that the open-source project works well and lives long.

[avr5500]

> A hosted version need not necessarily target maximum money making. You could run it as a nonprofit whose goal is to ensure that the open-source project works well and lives long.

That actually makes sense.

Also will apply to FOSS funds like - https://floss.fund

[nh2]

For an example of how it can go down otherwise:

papercups.io – Open-source alternative to Intercom

Funded by Y Combinator

https://web.archive.org/web/20230404011725/https://papercups...

https://news.ycombinator.com/item?id=26527268

https://news.ycombinator.com/item?id=24133719

I don't know why it shut down (my guess: didn't pan out with the typical revenue growth goals of a startup), but having a hosted version might save your project from such fate, making enough money to fund you, or somebody you hire who's excited about working on Free Software.

Some pricing ideas: Free for noncommercial with limited data retention (like the Chatwoot free tier), 3-5 $/month for noncommercial individuals (e.g. users to put it in their website), more $/month for commercial. Add some more expensive enterprise plan that supports Microsoft EntraID groups for externalised permission management and you're good to go :) Add easy DB import/export functionality, so people can switch between hosted and self-hosted. A lot of people will be happy to pay for the convenience of hosted. Host in EU for best data protection (makes it easier for people to sign up).

For billing, probably good to use a Merchant-of-Record service such as paddle.com (our startup likes it) to be able to sell internationally without having to deal with international taxes.

I wish you and the project lots of success!

[rancar2]

Chatwoot’s free hosted tier has limited data retention and only offers US-based data centers. However, its core functionality can be self-hosted anywhere under its MIT license.

For the OP: nh2’s advice above is worth considering. Additionally, I’m not sure Libredesk is mature enough to be a strong open-source alternative to Chatwoot yet—unless the user only needs email support. Chatwoot’s base installation, licensed under MIT, includes most of the features non-corporate users would want, such as multi-channel support, and generally offers more functionality than Libredesk under its AGPL license.

[Jolter]

Would hosting in Europe really help?

I’m under the impression that current EU case law makes it impossible for EU-based government entities to store personal data in services owned by US entities, no matter where they are hosted.

Maybe I misunderstood this?

Someone has told me that all the EU governments using Office 365 are basically violating the GDPR and getting away with it (for now). Any truth to that?

[yashasolutions]

Nice work. I cannot test the demo as it is down now but just wanted to share I have been using open source solution Zammad and I am pretty happy about it. It's fully open source (it's a ruby app - so different stack.) and pretty robust. The frontend is a bit oldish, but it has a lot of nice features. The single binary, easy deploy + fresh UI is interesting thought. Are you planning multilingual features?

[avr5500]

It's up now, yes planning to support multiple languages.

[d1sxeyes]

Looks great!

Mobile experience is not great but I think it’s sensible not to focus on that yet, the number of users doing customer support from a mobile phone is pretty small, even the big players in this space like ServiceNow etc have suboptimal mobile experiences.

Do you have any plans on how to monetise this or is this a labour of love?

[avr5500]

Mobile is not the focus, it can be done later as an app or something else. No plans to monetize; yes, it's a labor of love.

[d1sxeyes]

Lack of a monetisation strategy is a concern for folks who might worry that you end up running out of time/motivation to work on it, which might be a roadblock to wide adoption

That said, if you're only in it for the joy of creating something amazing, and don't really worry about whether you'll be able to keep maintaining it indefinitely, then keep it up!

[wiradikusuma]

The UI looks good! I can't seem to find the channels it supports. Email? X? Instagram (IG) comments?

A support desk is all about channel integrations, for example:

- replying IG comments from the app

- replying Telegram chats from the app

- replying [e-mail] from the app (SMTP & IMAP integration)

- basic CRM so that we that the customer doing all 3 above are the same, so the agent has "context"

- integration with external CRM

I hope I'm not offending anyone, but if any integrations are supported, you should emphasize this.

[chrismorgan]

From the screenshot, in an AI assistance dropdown on a reply form: “Add Empathy”. Ouch.

[ddtaylor]

Abolish corporate communication apathy, but don't point the finger at a dictionary as the reason it exists.

[darkwater]

The demo looks really really good! And by looking at commit history, looks like you are a solo developer behind this. Kudos to you! Out of curiosity, do you have plans to make it a business or just keep it as a "hobby project"? Something like this is going basically to be used only by companies, no hobby users (unless someone wants to maintain a support desk for family & friend about something they are an expert...)

[srameshc]

Nice name , looks great and very useful !! How you built a single binary of Go and Vue is very neat, I never thought about doing something like this.

[KingOfCoders]

This is also something I love about Go,

https://www.inkmi.com/blog/simplicity-of-golang-systemd-depl...

[srameshc]

thanks for sharing, there is always something to learn

[fguerraz]

This is great.

Sadly, apart from “looking good”, there seems to be no documentation, or visible roadmap. So I can’t evaluate if it’s worth looking into or not.

Keep up the good work!

[avr5500]

Thank you! Adding a roadmap to v1.

[mofirouz]

It would be wonderful if it could be used as a ticketing solution where the ticket could have severity. So customers could login to create a ticket as part of their org.

Also data sources from either third party in-house database or something HubSpot would be great!

[techn00]

I love single binary apps, easy to deploy.

[hamdouni]

It still needs postgres and redis

[remram]

Looks good! I'm currently using OsTicket (PHP) with a small team. Pretty happy with it but could use more automation. What I like best about OsTicket is that I was able to configure it completely transparent, there are no auto-messages sent to users and no reference numbers of boilerplate in outbound messages.

Happy to see more opensource options in this space, especially with a focus on customizable automation.

[dinu99]

Looks nice, congratulations. Can you add knowledgebase section? I'm looking for an alternative to freshdesk/zendesk. I would recommend monetizing this via hosted service like https://frappe.io/ does with ERPNext project.

[rushabh]

checkout https://frappe.io/helpdesk

[OldMatey]

Fantastic. This is far overdue and looks like a solid start. I have been particularly surprised by how few options there are in the CS space that are open source and options like Zendesk are SO expensive and the pricing tiers and cost structures are prohibitive for many small organizations.

[calvinmorrison]

Anyone using RT still here?

My idea tracking system is one that exists in email, but doesnt send a billion obnoxious "your ticket has been created" emails and butcher the contents. So far I've got it kind of working with RT as a no reply listener. We watch all comms but it doesnt reply to anything.

[gregoriol]

Interesting project!

However, installation documentation is quite strange: it says "requires postgres and redis to run" but there is zero other information. Is there any version requirements? any configuration file to set the hosts/auth for them?

Hope this helps, keep the good work!

[avr5500]

The Docker file specifies versions: https://github.com/abhinavxd/libredesk/blob/main/docker-comp...

I will include the versions in the installation docs for binary installations. Thanks for the feedback.

[yoeven]

Are there plans for future integrations into messaging platforms like Whatsapp or Telegram?

[ketan-10]

Looks awesome, Congratulations :)

[truepath_app]

I might be missing something obvious, but how does a customer file a ticket/request?

[avr5500]

Right now the only way to create a ticket/request is by adding an inbox, and sending an email to your configured email address.

[insane_dreamer]

Nicely done. Canned responses (the ability to create/manage/insert them) would be helpful.

Does it support email? That would be key. The demo looks more focused on chat.

[duckb]

Looks beautiful. How difficult would it be if I want to develop a plugin - Mood detection, Chat Helper (LLM linked), ...

[rgbrgb]

This is so cool, thanks for sharing. Can you talk about why you made it and why you made it free / open?

[nh2]

Looks great!

It seems to be email only so far. Do you plan to add chat, like Intercom and Chatwoot have it?

[avr5500]

yes, once Libredesk is stable.

[mouse_]

If the AI response feature promises something I cannot deliver, who is liable?

[rad_gruchalski]

You are.

[ddtaylor]

What method do we have to fund this or ensure it can work for our needs.

[oever]

Does it support the scenario where part of the team is only on IMAP?

[snvzz]

Mandatory Affero GPL warning.

Ensure you understand the license before deploying.

[mr-karan]

I'm curious to understand the implications of the AGPL-3.0 license in the context of this project. It sounds like the author doesn't want big corps to offer a hosted version of this tool (because doing that would require them to release the source code of any modifications they make to the software, which many companies are reluctant to do) and it's also an OSI approved license (unlike the recently famous SSPL or BSL). What's so wrong about AGPL-3.0 then?

The intention is often to prevent companies from building proprietary services on top of open source software and I feel AGPL 3.0 is a sensible choice here.

[rancar2]

Chatwoot has a lot more features with its MIT license.

[vishnumohandas]

Looks amazing, great demo!

nit: it'd be nice if we could resize the sections

[Aeolun]

Haven’t really seen anything else but it certainly loads fast!

[revskill]

Single app meanz j bundle all asset and code in a binary?

[qwertox]

This looks very promising. Does it have an API?

[avr5500]

Thanks,

Yes I will adding an API user support, you will be able to interact with Libredesk with APIs.

[mjoin]

Not in the domain but it looks great :) Congratulations

[crazymoka]

When it turns to beta, I'll be using this.

[registeredcorn]

It's a neat idea and I'd like to see how this evolves over time! Keep up the great work!! :)

I do have notes/feedback based off of what I see in the demo site:

# SLA

The SLA stuff under the Admin panel needs to be way, way more robust if I were to ever use it. Many of the clients I work with have different SLA expectations based on the severity of an issue. They also might work in different time zones from Support staff, so calendars may need to be set to "countdown" in relation to either client local time, or support local time, depending on what is decided in the contract. In addition to time zone stuff, there's also issues with what constitutes a holiday; we may have holidays that they don't, or vice versa. Unless something is covered under a niche 24/7/365 contract and is marked as Critical, we wouldn't want the "countdown" to trigger until the next business day. This matters a lot, because many US holidays are on Friday, so tickets will go unanswered for 3 full days.

Here's an example of multiple, different SLA requirements for one client:

* A Critical should have a response within 1 hour (24/7 coverage), and a resolution within 12 hours (24/7 coverage).

* A Low should have a response within 5 business days, and a resolution within 30 business days.

Now imagine that, but multiplied by many, many clients. Instead of having however many SLAs for each situation per client, we like to have "compact" SLAs where there are clauses for different severity ratings of a ticket, and factoring in a bunch of other stuff, for individual clients. Client 1 paid for 24/7 for criticals, but not other stuff, so Client 1 SLA is (this). Client 2 paid for 24/7 for all things, so Client 2 SLA is (that). Client 3 paid for the most basic coverage, so Client 3 SLA is (the other).

We would need an SLA system that accounts for various factors and verbiage within a ticket received. Sometimes it also depends on who files the ticket with us. Is it the "problem child", or is the CEO?

# Time zone (as it relates to SLAs)

In the contract, it is usually formally agreed that if no support staff lives in the clients local time zone, then the SLA response time will not start counting until it applies to the support staffs local time zone. Exceptions are made if they pay for 24/7/365 coverage. So, even if support receives two different "important" tickets from two different clients, one may take priority over the other because the service coverage they have paid for.

## Example(s):

* Client 1 files a ticket at 2PM their time; no 24/7 coverage. It's 4AM for local Support. Support's response time doesn't "start the countdown" until 8AM in Support's time zone.

* Client 2 files a ticket at 11AM and is supposed to have a response time within 4 hours; no 24/7 coverage. The ticket is filed on July 4th (Independence Day in the US) for the Support staff. The 4 hour "countdown" doesn't start until 8AM on July 7th. (3 day weekend)

* Client 3 files a ticket at 2PM on July 4th; have 24/7 coverage. It's 4AM for local Support on July 4th. The "countdown" has started.

# Holiday

It would also be nice if the option for "New Holiday" adjusted automatically each year. We have a fair number of clients, and manually updating Holiday's consistently (and accurately) is not something I see people doing particularly well at the start of each year. It's kind of one of those things you end up catching around September, and realize all of your metrics for the last few quarters have been off.

# FRD and RD text?

In the conversation with Felix Morin, I see warning labels for "FRD Overdue" and "RD Overdue". What are those triggered by? I didn't see anything in the admin area that seemed to point to it. It would be good if, on hover, it would give a brief explanation of what the trigger is. Or is it something that the person who filed the ticket did, somehow?

[avr5500]

Hey, thank you for taking time to write such a detailed feedback! Really appreciate this.

The automations are just for that, in automations admin tab you can create rules like email has text `@google.com`, `[e-mail]`. If true you can apply certain actions like set SLA, assign to a specific agent, etc.

You're right this SLA system is not flexible enough, I will work on improving this feature.

Regarding the "FRD Overdue" and "RD Overdue" labels, they show when the First Response Deadline or Resolution Deadline for a ticket has passed. Yes adding some hover text makes sense.

Maybe the https://libredesk.io/docs page needs a concepts section.

Again thank you for the feedback.

[registeredcorn]

Ah, nice. Thanks for the clarification! Really looking forward to the future of this system. :)

[TechDebtDevin]

Cool idea!

[curtisszmania]

We all know open-source customer support tools are still the wild west, and Libredesk might just be the sheriff in town.

Libredesk's single binary app approach is refreshing - simplicity can be beautiful when done right.

As someone who's been around the block a few times, I'm intrigued by the potential for self-hosted support desks to disrupt traditional SaaS models.

In fact, I remember when our own company was small and struggled with customer support. We ended up building something similar to Libredesk, but it took us months of trial and error - not exactly scalable.

So kudos to you, abhinavxd, for creating a tool that could potentially make all our lives easier (or harder, depending on how we choose to use it).

[fifticon]

this curtismania comment should be flagged as spam, but I can't do so from the interface. I don't know what he is eating, but it is not carbon-based. Possibly he has been eating marketing-copy, directly.