Pro AI · v3.0.764 · New
Your founder posture,
drawn from how you actually run the company.
Founder Insights is a weekly cross-recording read on six dimensions of operator posture — vision clarity, conviction, customer obsession, capital discipline, hiring posture, operating cadence. Each dimension surfaces a strength signal, the failure mode adjacent to it, and one Practice Move for the week. A mirror, not a grade.
Built for the operator running the company. Founder Insights is a working-document reflection tool for the founder whose own conversations are in the library. It is not a board diligence instrument, a fundraising-readiness scorecard, or a hiring tool for use against other people.
Why this reflection, and why now.
A founder's week is mostly conversations — investor updates, board calls, hiring loops, customer interviews, all-hands, 1:1s with the leadership team, the late-night call with a co-founder. What you say in those conversations is the most honest record of how you are actually running the company. It is more honest than the board deck, more granular than the OKRs, more current than the last 360.
Founder Insights exists because that record is most useful when it lands as a weekly six-dimension mirror rather than a 6,000-word retrospective or a forty-question self-assessment. The dimensions sit at the right altitude for the prompt — concrete enough to score against in your own words, broad enough to apply across a fundraise, a hiring sprint, a churn investigation, or a difficult board conversation. The framework was written for an operator running their own company in calendar weeks; it does not generalise to evaluating other founders, and we make that explicit.
This is not a personality framework, not a fundraising-readiness scorecard, and not a board diligence tool. It is a set of operator dimensions drawn from the existing canon — Paul Graham's essays on founders and conviction, Ben Horowitz on hard-thing decision-making, Peter Thiel on definite versus indefinite optimism, Sutton & Rao on scaling discipline — synthesised into six dimensions that an operator can hold in their head between Monday-morning planning and Friday-evening retro.
The six dimensions
Vision. Conviction. Customer. Capital. Hiring. Cadence.
Each dimension surfaces (a) a one-line read in your own words, anchored in specific quotes from your library, (b) the strength signal, (c) the failure mode adjacent to that strength, and (d) one Practice Move for the upcoming week. There is no composite score, no leaderboard, no comparison to other founders.
Dimension 1
Vision Clarity
How consistently you articulate the same picture of the company across audiences.
Strength: the same three sentences land for an engineer candidate, an enterprise prospect, and a Series B partner. Adjacent failure: rigidity — the same three sentences when the audience needs a different cut.
Dimension 2
Conviction
Willingness to hold a position against pushback when the underlying reasoning hasn't moved.
Strength: a clear answer on the hard call without hedging. Adjacent failure: stubbornness — holding the position after the reasoning has actually moved.
Dimension 3
Customer Obsession
Frequency and depth of customer-specific reasoning in everyday decisions.
Strength: named customers, named workflows, named edge cases drive product decisions. Adjacent failure: single-customer capture — building exclusively for the loudest customer in the library.
Dimension 4
Capital Discipline
Relationship between spend posture and runway-aware decision-making.
Strength: the next hire, the next vendor, the next experiment is sized against the burn picture. Adjacent failure: false thrift — under-investing in the load-bearing hire because the spreadsheet feels safer.
Dimension 5
Hiring Posture
How you evaluate, decide on, and onboard people in your own words.
Strength: specific bar, specific reasoning, specific risk acknowledged on every offer call. Adjacent failure: over-indexing on credentials, vibes, or the most recent reference call.
Dimension 6
Operating Cadence
How you sequence, prioritise, and follow through on commitments week to week.
Strength: last week's commitments visibly close out this week. Adjacent failure: calendar capture — the cadence is the calendar's, not yours.
No composite. No leaderboard.
There is no overall "founder score." The dimensions are independent reads, and treating them as a single number would imply a ranking the framework deliberately does not produce. The point is the mirror, not the grade.
How a dimension card reads in the app
Strength signal
"This is working."
A one-line read in your own words, anchored in a specific quote from this week's library.
Example: "You walked the Series B partner through the same three-sentence picture you walked the candidate through on Tuesday. That picture lands for both audiences without rewriting."
Adjacent failure mode
"Watch for this."
The failure mode that lives one step further along this dimension. Surfaced because the closer you are to the strength, the closer you are to the adjacent risk.
Example: "The same three sentences worked twice. They will not work for the Friday all-hands — that audience needs the operational cut, not the pitch cut."
Practice Move
"Try this this week."
One concrete move for the upcoming week. Specific, time-bounded, and addressable from inside your current calendar.
Example: "Draft the operational cut of the picture before Friday's all-hands. Three sentences. Different from the pitch cut. Open the meeting with it."
Safeguards on the card itself
Five guardrails, surfaced where you will see them.
1. No composite score. Each dimension reads independently. There is no overall "founder score," no leaderboard, no comparison to other founders. The framework deliberately refuses the ranking premise.
2. Strength always shown with adjacent failure. Every strength surfaces the failure mode one step further along the dimension. The closer you are to the strength, the more salient the adjacent risk — the card makes that explicit rather than letting the strength land alone.
3. Minimum sample of fifteen quotes per dimension. Below fifteen, a dimension shows a placeholder asking for more conversations before issuing any read. Insufficient sample is not a low score.
4. Reads are anchored in your own quotes. Every one-line read links to the specific moment in the library that drove it. The user is never expected to trust the bottom-line label without the receipt.
5. Methodology and limitations always one tap away. The card surfaces a link to this page from inside the app. The framework's scope — for the operator running the company, not for evaluating other operators — is stated on the card, not buried.
What Founder Insights is — and isn't — for.
It is for: a founder running their own company who wants a weekly mirror on the posture they are actually taking across the week. A solo founder using it as a working-document reflection between Monday planning and Friday retro. A two-founder team using it as the prompt for the standing co-founder 1:1. A founder-CEO using it as the prompt for the conversation they would already be having with their board chair or their coach.
It is not for: evaluating other founders, scoring portfolio companies, ranking candidates, grading acquisition targets, board diligence, or any other consequential decision about another person. The Colorado AI Act's high-risk classification for AI systems making consequential decisions in employment, and the EU AI Act's parallel framework, both apply to anyone using founder-posture-inference AI in those contexts. Don't use Founder Insights as an input to investment, hiring, or acquisition decisions about other people without your own legal review and your own ground-truth instruments.
Founder Insights is also not a coaching engagement, not a board self-assessment instrument, and not a clinical or therapeutic tool. The read is most useful as a prompt for the conversations you would already have with the human professionals around you.
Privacy & deletion.
Scoring runs server-side. All Founder Insights inference happens on Bonfiyah's backend. The iOS app renders the six dimension cards, the strength + failure-mode bands, the source quotes, and the Practice Moves — it does not run the LLM itself.
No new consent surface. The content fed to the LLM is whatever you have already accepted for AI processing under Bonfiyah's existing AI consent flow. No separate consent is required for Founder Insights.
Library scope is your own conversations. Founder Insights reads across the recordings in your account only. It is isolated from every other Bonfiyah user's library. The framework is built for the operator whose own voice is in the library; reading it across someone else's recordings would defeat the design.
Entries are deletable at any time. Delete a library entry and it is excluded from all future Founder Insights runs. Past dimension reads are not retroactively recomputed; you can wipe the entire Founder Insights history via Settings → Privacy → Delete Founder Insights history. Deleting your account cascades.
Off-switch in Settings. Settings → Notifications & AI exposes a Run + Notify toggle for Founder Insights. Run off disables the weekly inference entirely; Notify off keeps the inference but suppresses the morning push.
FAQ
What is Founder Insights?
A Pro AI cross-recording read on six dimensions of founder posture — Vision Clarity, Conviction, Customer Obsession, Capital Discipline, Hiring Posture, Operating Cadence. Each dimension surfaces a strength signal, the failure mode adjacent to it, and one Practice Move for the week. Refreshes weekly on Sunday at 06:30 UTC; manual refresh available.
Who is this for?
Founders running their own companies who want a weekly mirror on the posture they are actually taking. Built for the operator whose own voice is in the library — not for evaluating other founders, scoring portfolio companies, or grading candidates.
How does Bonfiyah avoid turning this into a personality verdict?
No composite score, no leaderboard, no comparison to other founders. Every strength surfaces the failure mode adjacent to it. Minimum fifteen quotes per dimension before any read renders. Every one-line read links to the specific moment in the library that drove it. The point is the mirror, not the grade.
Where does the inference run?
All scoring runs server-side on Bonfiyah's backend. The iOS app renders the cards; it does not run the LLM itself. No separate consent surface — the existing AI consent flow governs what content reaches the LLM. Settings → Notifications & AI exposes a Run + Notify toggle.
Could this be used to evaluate someone else's founder posture?
We strongly recommend against it. Founder Insights is built for the operator running the company. The Colorado AI Act (effective Feb 2026) and the EU AI Act both treat AI systems making consequential decisions in employment as high-risk. Don't use Founder Insights as an input to investment, hiring, or acquisition decisions about other people without your own legal review.
Can I delete my Founder Insights history?
Yes. Settings → Privacy → Delete Founder Insights history wipes every prior week's dimension reads, strengths, failure modes, and Practice Moves from your account and Bonfiyah's backend. Source library entries are untouched. Deleting your Bonfiyah account cascades.
Tell me when Founder Insights ships its next refinement
We're iterating the six dimensions, the strength + failure-mode pairing logic, and the weekly Practice Move generator. Subscribe if you want the changelog when those land.
No spam. We use ConvertKit. See our privacy policy.