You don’t need a Playbook
You need to know where the bodies will be buried in your business 10 years from now
Pixabay - This picture was the first thing that reminded me of pattern factories.
The Pattern Factory
TechBio founders don’t need playbooks.
They need a Pattern Factory. If that intrigues you - reply to this mail.
I’m finally building the thing I kept wishing someone else would build. The things that took me 10 years to learn.
If you spend enough time in TechBio — inside real conversations with founders, operators, regulators, and investors — you start to see the same design and anti-design patterns.
Some patterns come from music. My favorite is something I learned playing in jazz ensembles: “If you can only hear yourself, you’re playing too loud”.
Some come from engineering. The elephant in the room in enterprise software: BlindLeadingTheBlind.
And some come from the hard-earned scars of companies that almost made it. Trade-off or die.
It’s been 2.5 years since I started writing an essay on design patterns every Friday.
Today is issue #126.
Between the podcast and the newsletter, I’ve had 50+ 1:1 conversations with leaders across the industry.
And slowly, something clicked:
Different companies, different technologies — same underlying patterns.
And even more interesting: the same anti-patterns shows up again and again, quietly crashing into competitive positions.
Founder–market misfit
“Science and technology first, distribution later”
Outsourcing clinical data - your core IP
Scaling operations before product clarity
Underestimating process power
Overweighting founder intelligence
Hiring for prestige instead of performance
And on the positive side:
Counter-positioning against incumbents
Narrow wedges into large markets
Pattern reuse across indications
Distribution leverage through data network effects
Process power from robotics and automation
I realized something.
TechBio founders don’t need another playbook.
We need a Pattern Factory.
Not a checklist.
Not a “10 mistakes founders make.”
But a Pattern Factory that shows what actually works today — directly from real conversations, transcripts, and operator experience.
So… I’m building it.
Last year, I decided to lift up the hood on Agentic AI. I took a course with the outstanding Amir Feizpour (ai.science) and developed a working application using LangGraph to promote my latest book.
It was a lot of fun and I got to learn Python for the first time in my life.
I talk about my experiences over here in The Great Agentic AI hoax
I developed my own opinions on agents:
Not doing state management in Python code
Not forcing users into a DAG mental model
Naturally map to how business people AND engineers think.
I’m using a stateless message protocol for the agents where they exchange messages on a decision tree. The agent supervisor and humans talk with a put-and-take protocol - like a parent playing put and take with a baby.
Having a lot of fun.
Internally, the prototype already ingests:
podcast transcripts
Substack essays
pattern definitions
anti-pattern examples
founder quotes
relational links across companies
The result is something I haven’t seen before:
A Pattern Factory for TechBio — powered by people like you and me.
Not Elon Musk. Not business books. Not LinkedIn Gen AI content. Not retired drug company CEO. Not Taylor Offer.
People like you and me.
Next week I’m starting validation calls with founders and investors to shape where this should go.
If you want an early look, reply to this email with the word PATTERNS - tell me your favorite, nastiest anti-design pattern.
More to come — this is the most fun I’ve had building in a while.
This week on Life Sciences Today
Today on Life Sciences Today, we’re joined by Jeff Gombala, Founder and CEO of Evidexa — a startup tackling one of the biggest blockers in health-tech adoption: the evidence gap.
After 12 years in digital health, including as a partner at ZS Associates, Jeff is now building a company around a bold idea: digital twins of human behavior that let innovators run engagement and behavioral studies in silico instead of spending 16 months and $600K on real-world trials.
Evidexa helps prescription digital therapeutics, clinician workflow tools, and patient-facing apps break the “no data, no pilot” loop—and stand apart from traditional IT-checklist evaluators.
Let’s dive into Jeff’s journey and what Evidexa means for the future of evidence generation.
Watch the show here
About Me
I’m a 5X founder who learned hard lessons the hard way. Today, I help TechBio companies secure growth and maximize revenue with design patterns.
My private network is trusted by leaders from Medable, Menarini, Merck, IQVIA, Medtronic, Flatiron Health, Lindus Health, Debiopharm, Unlearn, Cornerstone AI — and 970+ more.
👉 Join here
If you love my writing — share it
If you want more like this — subscribe to Clear Thinking
right on