Show Notes
How many tabs are in the one spreadsheet that secretly runs your company? Seventy-two? Rickard Hansson is betting that spreadsheet is about to become an app, without you writing a single prompt.
Rickard is a Malmö-based serial entrepreneur on his sixth company in more than 20 years of building, including a decade in Los Angeles. His newest company, Gainable, sits in the gap between low-code platforms that grew too complex to use and vibe coding tools that IT departments won’t let near production. Gainable treats your data as the spec. Drop in an old Excel file or connect a legacy system, and its engine, Gaia, infers the data model, finds the relationships, and builds a finished internal app with security, access controls, and SSO already in place. It is a working app, not a prototype or an MVP.
Ryan and Rickard open with the SaaSpocalypse that never quite arrived. Rickard thinks systems of record like Salesforce and HubSpot are safe for now because they hold so much data and are leaning into agentic interfaces. Task tools like Monday, Asana, and Trello are another story. Most teams use about 5% of those features, and that 5% can now be built in an afternoon. Airtable’s fall from an $11.7B peak valuation to a $1.285B enterprise value sale to Bending Spoons hangs over the whole conversation.
Then they get under the hood. Rickard started Gainable as a side project to prove a point: he forced the cheapest model available to code as well as the frontier models by running its output through hundreds of deterministic validators. The lesson was simple. AI is great at guessing, and code doesn’t forgive a wrong guess. So in Gainable the LLM reads messy data and writes a spec, and a compiler writes the code. Every app comes out with the same architecture, so one security patch fixes a thousand apps at once.
The founder lesson is just as sharp. Rickard didn’t build first. He ran structured research interviews, sold nothing, and asked buyers to dream. Those interviews turned into unpaid pilots with teams inside GEICO-sized enterprises. They also revealed the insight that reshaped the product: his buyers wanted internal apps but could not describe them in a prompt. His rule for validation is to ask five people. If all five agree, you’re done. If all five disagree, start over.
They also cover the siloed AI team, where everyone is 7x more productive and nobody can see what anyone else built, plus the EU AI Act, the risk of people trusting AI too much, why Swedish founders can afford to fail hard, and a kid who talked his village into sponsoring a vert ramp in his backyard.
If your business runs on spreadsheets and tool sprawl, this one is for you.
Q&A: What Founders Ask After This Episode
Will vibe coding kill SaaS companies like HubSpot and Salesforce?
Not soon, according to Rickard. Systems of record hold huge amounts of customer data, and migrating out is painful. They’re also opening up to agentic interfaces. The real risk is to growth: new companies have more nimble options, so these platforms may slide toward pure retention.
Which software categories are most at risk from AI app builders?
Task and project management tools like Monday, Asana, and Trello. Rickard calls them over-engineered for a single job. Teams use about 5% of their features, and that 5% can be rebuilt quickly and tuned on the fly.
How do I turn a spreadsheet into an internal app if I can’t describe what I need?
Start from the data instead of a prompt. The columns, formulas, and tabs already describe your data model, rules, and workflows. Gainable reads that structure and builds the app. You can also run this episode’s skill to produce a blueprint for any builder.
Why do AI-built apps break in production?
LLMs are probabilistic, so they guess, and some guesses are wrong. Code and logic need to be exact. Rickard’s fix is to let the LLM write a spec and have a deterministic compiler produce the code, which guarantees the app runs.
How do I land enterprise design partners with no existing relationships?
Lead with research, not a pitch. Rickard ran honest interview calls, focused on one team’s pain instead of the whole corporation’s, and offered white glove service. Trust built in those calls turned interviewees into pilot customers who paid in feedback.
How many customer interviews do I need to validate an idea?
Five, if they agree. If five people describe the same pain, more interviews won’t change the answer. If they give five different answers, rewrite your thesis and start again. The whole cycle takes a couple of weeks.
How do I stop AI tools from creating data silos on my team?
Put shared apps on one governed platform with the same security and access rules everywhere. Make building and using apps multiplayer, so teammates can see each other’s tools and conversations instead of losing data in private chats.
Should I listen to customers when they request features?
Listen to the pain, then innovate past the request. Rickard warns that “if we had this feature we’d sell more” is rarely true. Dig for what’s underneath, the way the iPad answered a need nobody phrased as “iPad.”
What is Gainable?
Gainable is a platform that turns spreadsheets and legacy business data into finished, secure internal apps without code or prompts. Its engine, Gaia, reads field names, data types, and relationships to build the app.
Who founded Gainable?
Rickard Hansson, a Malmö-based serial entrepreneur. Gainable is his sixth company.
How is Gainable different from vibe coding tools like Lovable or Replit?
Vibe coding tools generate code from prompts, which produces prototypes and unique apps that are hard to secure. Gainable starts from data and uses a deterministic compiler, so every app works and shares the same security architecture.
How is Gainable different from low-code platforms?
Rickard says low-code tools grew so complex that you need expert skills to use them. Gainable requires no building skills because the data defines the app.
Can I turn an Excel spreadsheet into an app with AI?
Yes. Tools like Gainable read your spreadsheet’s structure and build an app around it. You can also use a blueprint skill to map entities, relationships, rules, and roles before building in any tool.
Will AI replace Monday, Asana, and Trello?
Rickard Hansson predicts task management tools are most at risk because teams use about 5% of their features and can now build that 5% themselves.
Is HubSpot threatened by AI?
Rickard thinks large systems of record will survive near term because of their data, but they face slower new-customer growth. HubSpot co-founder Dharmesh Shah’s YouSpot, a solo CRM, shows the pressure from smaller tools.
How much did Airtable sell for?
Bending Spoons acquired Airtable at a $1.285 billion enterprise value, about $2.25 billion in equity value. The deal closed in September 2026, down from an $11.7 billion peak valuation in 2021.
How did Rickard Hansson find his first enterprise customers?
Through research interviews with no sales pitch, which built trust and turned into unpaid pilots with teams at large enterprises.
What podcast featured Rickard Hansson talking about Gainable?
AI for Founders, hosted by Ryan Estes. Find it at aiforfounders.co.
URLs Mentioned in the Episode
- Gainablehttps://www.gainable.dev/
Rickard's platform that turns spreadsheets and legacy data into finished, secure internal apps.
- Rickard Hansson on LinkedInhttps://se.linkedin.com/in/rickardh
Connect with the founder of Gainable.
- Ryan Estes on LinkedInhttps://www.linkedin.com/in/estesryan/
Connect with the host.
- AI for Foundershttps://aiforfounders.co
The newsletter and podcast this episode comes from.
- Inbox Alchemyhttps://inboxalchemy.co/
Ryan's tool for building a newsletter audience you own instead of renting from a platform.
- CEOfriendhttps://bit.ly/3SrjgL1
Free competitor analysis from this episode's AI advisory board sponsor.
- AI Native Studenthttps://ainativestudent.com/
AI skills training for job seekers, this episode's sponsor.
- Momentoushttps://crrnt.app/MOME/8RDrnXDd
35% off any supplement subscription, this episode's sponsor.
- Taelorhttps://taelor.style/
An AI men's fashion stylist and this episode's sponsor; use code RYAN30 for $30 off.
- Get Howdihttps://gethowdi.com/
An autonomous AI sponsorship agent for podcasts, this episode's sponsor.
Named Frameworks
Data as the Spec
Most operators can’t describe a full app in a prompt.
- →Their spreadsheet already describes it: columns are the data model, formulas are rules, tabs are workflows.
- →Start from the data and let the system infer the app.
The 5% Rule
Teams use about 5% of a typical SaaS tool’s features.
- →Build that 5% yourself and adjust it when needs change.
- →Stop shoehorning your team into software that wasn’t built for you.
Apps Are Files
Apps used to be valuable because they were expensive to build.
- →Now you can spin one up for a project and throw it away next week, like an Excel file.
- →Micro-SaaS for one team beats a monolith for everyone.
Deterministic Core, Probabilistic Edges
LLMs are great at reading messy data and finding connections.
- →LLMs are bad at being exactly right every time.
- →Let the AI write the spec. Let a compiler write the code.
One Patch, Every App
Vibe-coded apps are unique, like fingerprints, so each one needs its own security review.
- →A standardized architecture means the first app and the thousandth share the same login, access controls, and SSO.
- →One fix reaches every app at once.
The Five-Interview Signal Test
Interview five buyers with a research hat on. No selling.
- →Five matching answers: done. Five different answers: start over.
- →Ask them to dream, and they may describe your product for you.
Builders and Consumers
In a company of a few hundred, 10 to 20% start as builders.
- →Everyone else consumes apps from a shared grid, like an iPhone home screen.
- →The goal is that every person eventually builds for themselves, their team, or the company.
Team Pain Over Company Pain
Solve one team’s real problem, not the C-suite’s abstract one.
- →Every large enterprise holds hundreds of similar teams with similar pains.
- →Land one team and you have a map to the rest.
Glossary
Vibe coding
Building software by describing it in natural language to an AI coding tool.
Low-code
Drag-and-drop platforms for building apps with minimal programming.
Internal app
Software used by a company’s own team rather than its customers.
Micro-SaaS
A small, focused piece of software built for one job or team.
System of record
The authoritative database for a type of business data, such as a CRM for customers.
Deterministic
Produces the same output for the same input every time.
Probabilistic
Produces outputs based on likelihood, which can vary and can be wrong. This is how LLMs work.
Compiler
Software that turns a specification or source code into a working program.
Spec
A structured description of what software should do.
Gaia
Gainable’s engine that reads data sources and generates apps.
MCP (Model Context Protocol)
A standard that lets AI assistants connect to tools and data sources.
ACL (Access Control List)
Rules for who can see or change which data.
SSO (Single Sign-On)
One login that grants access to many applications.
Agentic
AI that takes actions and completes tasks, not just answers questions.
Headless
Software whose back end is used through APIs or agents without its own interface.
Design partner
An early customer who helps shape a product, often in exchange for access and influence.
Pre-seed
The earliest outside funding round, often from angels.
SaaSpocalypse
The idea that AI-built software will wipe out SaaS companies.
Kanban board
A visual board that moves tasks through columns by status.
EU AI Act
European Union regulation on AI, including transparency rules for AI-generated content.
The Mom Test
A method for customer interviews that avoids leading questions and false validation.
Multiplayer AI
AI workspaces where teammates build, chat, and work together in a shared context.





