Date: 05.11.2025 (2025-11-05)
Timon Oberholzer (05.11.2025) Message: • Ich bin grade in der Gründung und arbeite grade an einer Plattform für Geschäftsprozessautomatisierung (ähnlich wie UiPath oder n8n), die auf zwei zentralen Innovationen basiert: • (1) einer neuen Generation von LLM-basierter Integration für Legacy-Software – selbst ohne API; und • (2) einem KI-Co-Piloten, der automatisch komplette End-to-End-Workflows erstellt – egal ob sie in einem Walkthrough-Video gezeigt, in Text beschrieben oder per Klick-Demonstration aufgenommen wurden.
Questions: • Context: team setup? internal and external? • Notes: • Partner at Cave -> Software contracting • UBS is customer - they have legacy systems • Two completely different worlds • He uses new Notion functions (Notion AI 3.0) • Chat panel incl. MCP API - CoPilot -> can trigger all features that Notion can do • Does not automate processes in general but makes his work much faster • Pitch deck drafting -> much faster • E.g. market analysis -> store to page -> correct manually -> then start new thread -> then create pitch deck • Can create databases as well -> you can plan a roadmap • Check Notion AI -> user experience is the best • Check app function • They used LangChain • Bot that reads order from mail • Worked well because they knew how the process should look like • Challenge to implement at organisational level - to much pre know how to build it • People have time • He looked at n8n • Difficult if process is complex / unclear -> human is needed • He thinks next step is Copilot - like the one for n8n • Has to get much easier • He thinks hands off flows that can be hands off are limited • A lot of them are automated in a deterministic flow • Exciting is the close collaboration with the human • POV corporate: for every agent you need a human counterpart that works with it - like working with a team member • Q: how can you use the knowledge of the human • When building n8n workflows • Steps are always the same - e.g. authentication • Transform the “sparring partner & junior worker” logic to workflows • Customer side: • They do not understand nothing • They try to catch up with digitalisation -> very political • UBS realised it • Most companies want to build inhouse atm. • UBS knows that they have to do sth but do not know how -> paralyse with Misstrauen • Many are open to talk but no exact need • Data privacy is very important -> a lot of Misstrauen when it comes to AI and automation • They over or underestimate capability of LLMs • Finance & Accounting feedback: • They had contacts with KMUs • It is difficult to get in -> the small do not see the problem not enough • The big ones -> data privacy • He does not know anyone top of mind • Treuhänder could be a good GTM • What do they do? • He is not anymore with Skincare • He is looking for new things -> document automation • Data to document engine (bootstrapped) -> last mile via editor • He wants to scale the business • Core team: 4 people • https://uhuu.io/en • First customers: personal network • Partner with Cave -> legacy business -> he automated a lot -> not a lot of time • How to help? • B2B SaaS people -> to share experience • Do they already have automations? Which ones? How well do they work? What about other departments? • What about legacy software? How do they solve that problem? • Biggest objections / challenges in his daily life? • Which tasks are the biggest pain? With magic wand -> what would he solve? • Which tasks would he focus on? • Which companies would he target? • General feedback to pitch?