About us
At Taskbase we’re lifelong learners

Our story
Where we started
Taskbase started as an ETH Zürich spin-off. Samuel Portmann, Daniel Niederberger and Jost Joller founded it in the field of adaptive learning: tools that help people learn in the most effective way possible.
What we believed in
We built the first versions of AI tutors. An AI tutor understands a learner well enough to guide them through the material, teach it and set exercises. It learns where the learner stands, what they are good at and where they struggle, what their personality and learning style are. In a classroom you are one out of 20, and no teacher can tailor the lesson to each student. An AI tutor can. That was the first thing we believed in: every person’s journey should fit their needs and their way of learning.
Where sales came in
Then Patrick Trümpi joined, and we merged everything we knew about learning and coaching with his experience in sales. Sales reps are learners too, and almost nobody was helping them improve day to day. That is how Revenue AI came about.


Our visionEmpower humanity through AI.
Timeline
From learning platform to Revenue AI
Eleven years, one idea: help people learn the way that works for them.
2015
A spin-off from ETH Zürich
Samuel Portmann, Daniel Niederberger and Jost Joller found Taskbase in adaptive learning: tools that help people learn in the most effective way possible.




2017
We win the tender of the Canton of St. Gallen
The canton wants one big learning platform. A small team takes the job on and spends the next years building Lernnavi.
Canton of St. Gallen
2020
The first hires
Sonja Heim, now COO, and Kristian Brünn, now CTO, are among the first people to join. In the following year the team keeps growing and the first AI tutors take shape.


2021
Lernnavi goes live
On 1 August 2021 the platform opens for German and mathematics at Gymnasium and Fachmittelschule. Learners get tasks that fit their level and feedback on every task. Klett and a few more clients follow.
Klett

2024
Patrick joins, and investors back us
Patrick Trümpi brings his sales experience. Mediahuis Ventures and Acrobator Ventures lead our seed round in June, and the team grows to about 30 people. By the end of the year around 30 customers use Taskbase for learning.
Mediahuis VenturesAcrobator Ventures

Jun 2025
One week, one decision
At an off-site in our chalet in Leysin we combine Patrick’s sales knowledge, which he shares on Substack and LinkedIn, with everything our first AI tutors taught us. The idea for an AI sales coach is born.

Sep 2025
The first customers of our AI sales coach
Crewmeister and Pix4D are the first to work with the AI sales coach. Taskbase already serves around 30 customers with its learning products by then, but this is a new product.
Oct 2025
We acquire Micromate
Micromate (micromate.ai) has experience in enterprise security and in chat integrations such as Microsoft Teams. Four new colleagues join Taskbase, two of them the founders of Micromate. Two more founders in the team raise the entrepreneurial spirit of Taskbase even further.
Micromate4 new colleagues
Christoph SüessLead Product Owner
Stefan SchöbLead Software Engineer
Jan 2026
The AI sales coach launches
We launch the product, and it is not good enough yet. We keep learning from how reps and leaders actually use it.
TTaskbase10:42✅ WHAT YOU DID WELL
📌 You ranked the incumbent vendor fairly
Skill: Competitive Positioning
You acknowledged the incumbent respectfully and explained the differences in usage, skill development and the living playbook. That strengthened your credibility with Marc.
⚡ BIGGEST LEVER · 🎯 one of your 3 focus areas
📌 Accept his priority before you open a new use case
Skill: Active Listening & Shared Understanding
The moment in your call:
“Okay, but we haven’t talked about the security use case.”
How you could handle it better:
“Understood. Before I open another use case: what is really top of mind in your first 90 days, and should I get back to you only after that?”
Why it matters: Marc had just made clear the topic is not top of mind. Changing the subject straight away made you sound pushier and stopped you from narrowing down his real first priority together. [Playbook: Qualification Framework]
📈 PROGRESS SINCE LAST TIME
On 6 October, picking up the customer’s answer precisely was already your coaching lever. Today the pattern showed again when, after Marc’s prioritisation, you opened a new use case straight away.
🌟 Keep your fair competitor positioning. At the next clear priority signal, first pause for a short reflection and one focusing question.
May 2026
Relaunched on our own AI agent platform
With the AI agent platform as its foundation, the product changes completely, and so does how our clients use it. The living playbook and the automations around the coach become quick to build.
Every Monday at 9:00, send me everyone in my territory who bought a licence last week and fits our ICP. Find three contacts per company, with contact data.
Got it. Your licences sit in your own product database. Shall I read them through its MCP?
Yes, use the MCP. My territory is DACH.
Done, built in 7 minutes. I’ll run it every Monday at 09:00 and post the result here. This week 4 companies would qualify.
New skillWeekly licence buyersMondays 09:00Territory: DACHICP from your playbookSource: your product database
Make it Tuesdays, and only companies with more than 200 employees.
Updated. Next run: Tuesday 09:00, companies over 200 employees.
Today
Revenue AI
More than 50 clients use Revenue AI: a playbook that updates itself, coaching on every call and the whole job around the calls automated. The rest of our business keeps its solid revenue and keeps growing. Our vision: empower humanity through AI.

A note from Patrick
Why we built a new Revenue AI
I am the sales expert at Taskbase. This is what I learned, and why we build the way we do.
I lived the problem
I have led sales four times, as Chief Sales Officer and VP of Sales, and I co-founded a company before that. Every time I saw the same gap: nobody had solved how sales reps get coached and really improve, day after day.
My first attempt was at Unique AI, where I was Chief Sales Officer. We built what every recording and revenue AI tool builds: we recorded calls and gave feedback on them. It never really worked.
Feedback on a call is not coaching
Our clients did not digest it. Only a very small share of reps did, mostly people who had just started and wanted to learn. Three things to improve and three done well is a snapshot of one call. It ignores how far the deal is, and it ignores the progress the rep has already made.
There was no cumulative score either. Nobody could see how a rep develops over time, and that is a huge motivational factor.
And the feedback rested on a few battle cards, MEDDICC, SPICED or SPIN, never on a complete playbook. A playbook is how I, as a sales leader, define how we sell: how we run discovery calls and demos, which feature we show for which problem, our unique value propositions, what sets us apart from each competitor, who our competitors are and how we handle objections. That is far more than a methodology and a few battle cards.
Nobody reads the playbook
Many sales leaders have no playbook at all. I always wrote one, and nobody ever read it. I got tired of changing, adapting and updating it, because it was a huge effort with almost no impact.
What we learned
First: AI can coach reps far better than simple feedback on a call, but it needs the context of the playbook. Second: we can build a playbook in a few hours from all past data, instead of a sales leader writing a huge document for months. It is good enough to coach reps and to get them through the day.
Third: learning fits into sales much better when you also automate the rep’s day. Using Taskbase in our own team, we saw that 90% of our work is meetings and cold calls. The other 10% is checking the AI: whether the follow-up emails are good, whether the day is prioritised well, whether the call list is right, whether the mutual action plans, the reporting and the forecast hold up. So we automate the whole job around the calls, and we coach reps on the calls in a new way, so that their skills improve.
An edge, built AI-first
At Unique we were always running after Gong. They had more engineers, more clients and more feedback. Our edge was European values and data protection, not features. Building Taskbase AI-first, we can build the features the big tools have in a short time, and we can individualise them for every company and even for every rep, while keeping a standard that is good enough for sales leaders.
That is the dichotomy every sales leader knows. You want to standardise, so you have a playbook that scales. But reps only use your tools when those tools fit them a little. Seeing that we can have both was our big aha in June 2025, and it is why we built a new Revenue AI.
Where this goes
Reps will always have a job and will always do calls. In mid-market and enterprise sales, human-to-human conversations will not disappear in any future I can foresee. But some buyers will buy only from AI, perhaps Gen Z. If you deploy AI salespeople one day, the only thing you will need is an up-to-date sales playbook, so that your AI knows exactly what to do at each step. The playbook is the ground truth.
Taskbase helps reps get through the day and make their best calls, and it keeps the playbook current from calls and best practices, so nobody has to rewrite it again. If you believe some buyers will want to buy without any human, you will be ready with us.
Our vision is to empower humanity through AI. Our mission starts with helping sales teams overachieve their targets.
Meet our team
The dedicated builders and thinkers at Taskbase.
Samuel PortmannCEO, Co-founder
LinkedIn
Sonja HeimCOO, Co-founder
LinkedIn
Kristian BrünnCTO, Co-founder
LinkedIn
Patrick TrümpiCRO, Co-founder
LinkedIn
Jetro CapiaghiCFO
LinkedIn
Daniel NiederbergerSoftware Engineer, Co-founder
LinkedIn
Jost JollerChief Architect, Co-founder
LinkedIn
Stefano MalagòSoftware Engineer
LinkedIn
Tanja MüllerLearning Expert
LinkedIn
Julia ZöllnerCustomer Success Manager
LinkedIn
Sebastian ZumbühlSenior Sales Consultant
LinkedIn
Daniela Gerber WongSales Consultant
LinkedIn
Hamza HajrySales Consultant
LinkedIn
Lukas AbeggSoftware Engineer
LinkedIn
Maja EffenbergSoftware Engineer
LinkedIn
Christoph SüessLead Product Owner
LinkedIn
Stefan NiederhauserProject Manager
LinkedIn
David KlaperSenior Software & NLP Engineer
LinkedIn
Stefan SchöbLead Software Engineer
LinkedIn
Yann WallisSoftware Engineer
LinkedIn
Jo HelmuthData Scientist
LinkedIn
Looking to join Taskbase?
We look forward to learning more about you and sharing what we’re building.
