The assistant that
works across your apps
Ask in plain language. Reda reads your channels, logs your hours, manages your projects and plans your meetings, without you switching apps. Connect your own ChatGPT or Claude account.
Reda docked beside Alma's calendar. The apps that support it share the same panel.
One assistant, inside your apps
Reda isn't a chatbot bolted onto one tool. It works across Snacka, Planera, Tida, Alma and Tempa as you, with your own permissions, so a single sentence turns into real work in the right app. Synas, the page builder, is coming next.
Close out your day
in a sentence
“Log 6 hours on the website rebuild.” Reda writes it to Tida and hands you your weekly total. No timesheet grid, no context switch.
Start a project
by asking for it
New client, new project. Reda adds it to the workspace's shared projects with the hours you name, so it shows up in Planera for planning and in Tida for time tracking at once.
Re-scope on
the fly
Plans change. Tell Reda a project is running long and it changes the weekly hours on the shared project, which Planera plans against and Tida reads, so your capacity view stays honest without a single form.
Agents
A bot with a mission,
and a person who approves
An admin gives an agent a name, a job description and the tools it works with: a mailbox it watches and a Snacka channel it posts in. When mail arrives, the agent posts a summary and a draft reply in the channel. A colleague approves it by replying with the approval word (for now in Swedish, godkänn), and only then is the reply sent.
Agents run on an AI key an admin adds for the workspace, the one place Reda does not use a person's own account.
Everywhere you work
Right where the question comes up
Reda has its own chat, and it also sits docked beside your work in Alma, Planera, Tida and Tempa, so you can ask without leaving the page you are on. In Snacka it answers in a thread.
Your model, your account
Bring your own model
Each person connects their own Claude subscription or ChatGPT account. No markup on tokens and no new AI vendor: prompts go straight from your server to the provider you already use.
€17 / mo flat
One price per workspace, for everyone in it, each on the account they already have.
What leaves your server
Reda runs on your VPS and keeps your chats there. The model it uses gets no database credentials and reaches your data only through the apps, as you. What you ask, and whatever it reads to answer, goes to Anthropic or OpenAI under your own account, never through Nordkastell. Both are US companies, and whether they train on it depends on your plan and settings with them.
Who sees your chats
A chat belongs to the person who started it. An admin can see in History that a conversation happened, where, and what the agent did, but not the prompts, replies or tool output.
Managed AI, coming soon
A fully managed workspace model with no key to bring, for teams that want it turnkey.
Put an assistant
inside your workspace
Reda comes with the AI app. Tell us about your team and we will set it up in your workspace.