5 teams, 24 hours, we ended up with five working prototypes shipping this week.
Tools
AI tool directory for teams that build instead of listening.
Every tool we use to build prototypes in a Corporathon: models, coding agents, automation, data and CRM. With a view on effort, cost and fit.
Built for
Tools in the Corporathon stack
OpenAI
OpenAI provides the model layer behind most prototypes built in a Corporathon: chat assistants, text processing, document analysis and function calls into your own systems.
LLMs & assistants
Claude
Claude is strong on long documents, structured analysis and traceable answers. Teams use it when quality and reasoning matter more than raw speed.
LLMs & assistants
Claude Code
Claude Code works directly in the repository: reads files, edits code, runs commands. For teams with developers it is the fastest path from idea to running prototype.
AI coding
Gemini
Gemini is the obvious choice when documents, mail and sheets already live in Google Workspace. Multimodal input makes it useful for image and video material.
LLMs & assistantsWorkspace
GitHub Copilot
Copilot speeds up existing development work without changing the workflow. For teams with an existing codebase it is the lowest-risk entry into AI-assisted development.
AI coding
Codex
Codex takes on scoped development tasks independently and returns proposed changes. Useful when many small tasks should run in parallel.
AI coding
Lovable
Lovable builds complete web applications from descriptions, including database and deployment. Business teams without a development background reach a real prototype instead of slides.
AI coding
n8n
n8n connects systems, AI models and custom scripts in a visual workflow. Self-hosting makes it the preferred choice when data must stay inside the company.
Automation
Make
Make maps processes as visual scenarios and stays manageable without an IT team. Well suited to processes with many branches.
Automation
Zapier
Zapier is the fastest way to make two systems talk to each other. For simple chains it beats any custom build.
Automation
Airtable
Airtable is the pragmatic data layer for prototypes: tables, views, automations and interfaces in one place.
Data & enrichmentWorkspace
Clay
Clay combines many data sources with AI research in one table. Instead of buying lists you build your own verified data base.
Data & enrichmentOutbound
Attio
Attio models relationships the way your business actually works instead of forcing you into a rigid object model.
CRM
HubSpot
HubSpot covers marketing, sales and service together. For AI projects it is usually both the data source and the target system.
CRM
Salesforce
In many large companies Salesforce is the binding source of truth. AI prototypes have to align with it, not the other way round.
CRM
Pipedrive
Pipedrive deliberately stays simple. For smaller sales teams that is an advantage because AI additions show impact immediately.
CRM
Microsoft 365
Teams, SharePoint and Outlook are where AI results must arrive in many companies if they are to be used at all.
Workspace
Microsoft Dynamics 365
Dynamics connects sales, service and operational processes. For companies on a Microsoft stack it is the natural connection point.
CRM
Instantly
Instantly handles sending, deliverability and campaign control. Quality comes from the data upstream, not from the sending tool.
Outbound
Lemlist
Lemlist combines email, LinkedIn and calls in one sequence and focuses on personalisation over volume.
Outbound
Free Corporathon tools
Three tools of our own with transparent logic and no registration.
Customer voices
What participants publicly say about the hackathon
Summarised takeaways from approved LinkedIn posts by teams that built with us.
From “I should really learn AI” to a shipped product, with zero coding experience.
No months of prep needed: a team, a bit of chaos, and it turned into a working product.
5 teams, 24 hours, we ended up with five working prototypes shipping this week.
From “I should really learn AI” to a shipped product, with zero coding experience.
After day one, people who had never opened a terminal were building in Claude Code.
OW
Oke Wilhelm
NavVis
The biggest effect wasn’t the tooling, it was the mindset shift: “I can build this myself.”
NT
NavVis Team
Marketing & ops
After day one, people who had never opened a terminal were building in Claude Code.
OW
Oke Wilhelm
NavVis
The biggest effect wasn’t the tooling, it was the mindset shift: “I can build this myself.”
NT
NavVis Team
Marketing & ops
Next step
Ready for your Corporathon?
Book a discovery call directly, we clarify goal, format and date in the call.
The calendar loads external content from Cal.com. Click to load it, or accept "functionality" in the cookie settings so it loads automatically next time.
FAQ
Frequently asked questions
By use case, data access and who has to operate the result afterwards. Not by vendor partnership.
No. Most teams ship a first prototype with two or three of them.
No. Prices are indicative for orientation. Always verify on the vendor page.
Yes. We work inside your systems whenever access is available.
You do. Code, workflows and prompts stay with your company.
We agree scope, data sources and access rules before the hackathon and only use approved material.
No. Licences are bought directly from vendors; we help with selection and setup.
Yes, for tools like n8n. That changes effort and operating responsibility, which we plan for.
One workshop day is enough for a working prototype in most cases.
You get documentation on setup, cost and owners so the go-live decision is solid.