This is a real Cadence workspace.

No login, nothing to set up. Everything below is live data from a seeded demo workspace: a real teardown, a real decision ledger, a real mission trace. You cannot break anything, so look around.

Today, in Explore workspace

What Cadence is watching right now.

9open opportunities

11decisions on record

30missions in flight

A real teardown

Bank-link drop-off at activation

reviseICE 8.0

Significant drop-off (38%) at bank-linking is a clear problem. However, the proposal lacks a defined target user, a specific testable hypothesis (no solution proposed), and detailed root-cause analysis for the 'confusing connection.' Needs substantial development.

Risks

Untargeted solution: Without a defined target user, proposed solutions may not address specific needs of the affected cohort, leading to ineffective changes.
Scope creep/misdirection: Lacking a specific solution or deeper root cause analysis, efforts could be misdirected or become too broad.
Confidence in solution delivery: High 'Confidence' score (9) is premature without any proposed intervention, indicating over-optimism about finding an effective solution.

What you cannot prove yet

Target user definition: Who are these 38%? What are their demographics, tech savviness, acquisition source?
Root cause analysis: What specifically makes the connection 'failed or confusing'? Is it specific banks, error messages, Plaid's UI, our integration, device issues? Qualitative data is needed.
Proposed solution/intervention: The hypothesis is not testable without a specific action or change being proposed.

Read a full public teardown, no signup→

The ledger

Every call, on the record.

Price Prism Plus at $7.99/mo with a family tier to followapproved

Trial-to-paid is a timing problem, not a price problem (interviews). Hold $7.99 and fix the paywall timing first; add a $12.99 family tier once shared goals ship, based on the design-partner willingness-to-pay.

Prioritize · Jun 15

Supersede launch fraud policy with per-user precision scoringstanding

The aggressive launch policy became the top churn driver (4.1% false positives, 2 declines tends to churn). Per-user baseline scoring plus a 1-tap confirm for medium-risk preserves the fraud catch rate while cutting false positives below 1%. Supersedes the launch policy (2026-02).

Prioritize · Jun 6

Supersede SQL-first with a natural-language question boxstanding

The SQL-first path gated out non-analysts, the majority of trial owners. A natural-language question box that compiles to warehouse SQL keeps the analyst power while opening answers to PMs and founders. Supersedes the SQL-first decision.

Prioritize · Jun 5

Ship round-up savings goalsstanding

Round-up sweeps directly serve the north star (funded goals in 30 days) by pulling savings from spend the user does not feel. Strong emotional pull in interviews, moderate build cost.

Draft · May 22

Kill the reverse-ETL parity betstanding

Critic matched the workspace parity precedent: parity bets that copy a commodity feature have not retained our ICP (the crypto-wallet kill on the other product is the same pattern). Reverse-ETL is differentiator-neutral for a warehouse-native analytics wedge. Kill; revisit only with real churn evidence.

Challenge · May 22

The whole trust ledger, wins and misses→

One mission, in motion

Add a TODO.md to repo root with one line of text

running

DraftGenerate a PRD for adding a TODO.md file at the repository root with the content 'Nothing pending right now.'done

EngineerCreate a new file named TODO.md at the repository root with the content 'Nothing pending right now.'. Stage this file and create a pull request.done

ReviewVerify the pull request created by the builder. Ensure the TODO.md file is in the correct location and contains the specified content. Confirm CI passes.dispatched

AnnounceFinalize the pull request and initiate the merge to the main branch.planned

Tear down your own pet feature

the beta is open for sign-ups