iBrokerage — first consumer product
Consumer finance direction on One, Kai, and PCHP; vault-based Plaid source and dated Aha Moment acceptance target.
TL;DR: iBrokerage is the consumer finance product direction on top of One, Kai, and PCHP: one consented import path, one trustworthy financial picture, one useful next number in about a minute.
Product plan as of 2026-05-19: the Aha Moment and sprint sequence below are dated targets.
Implementation review as of 2026-09-23: the current source uses a vault-backed Plaid path. Environment migration and legacy Plaid Item cleanup require separate proof.
Relations
- One — the relationship layer iBrokerage sits under.
- Kai — finance specialist that assembles portfolio/import context and savings signals.
- Nav — privacy and consent guardian direction for account-read review.
- PCHP — every Gmail, brokerage, bank, or provider read must be scoped and auditable.
- BYOA — model-agnostic by commitment.
- Aha Moment — the canonical onboarding gate and Sprint 3 acceptance criterion.
- Personal Operating Layer — the strategic position iBrokerage is the v1 launch surface for.
- MLX on One surfaces — on-device inference direction for fast local parsing on Apple silicon.
The Aha Moment
A consumer-product test: can a brand-new user get to "this is worth my time" in under one minute?
iBrokerage's answer: connect an approved source, let Kai assemble scoped finance context, and surface a dollar number — fees the user is paying, idle cash, duplicate subscriptions, missed yield, or another defensible signal. The savings number is dollar-denominated because that is the only number consumers reliably remember.
This shape — one connect, one number, one screen — is the Apple frame in financial software. See Apple frame.
The north-star user persona
Not a real person. The composite north-star user used in every iBrokerage product decision:
- Senior tech executive, mid-40s, time-starved.
- $4M+ portfolio across 8 accounts (Schwab, Fidelity, Robinhood, 401k, employer RSUs, HSA, joint household, kids' 529).
- Tracks nothing in one place. Pays fees he doesn't see. Has cash earning nothing he hasn't reconciled.
- Will pay for a product that gives him 90% of a CFO's clarity in 5 minutes a week.
If a feature does not visibly help the north-star user inside one sitting, it does not ship for v1.
Implementation source and deployment boundary
The current Plaid source provides a read-only vault flow for connection, exchange, snapshot refresh, relink, and removal. The backend handles Plaid tokens and readable responses transiently; the device seals the returned access token and financial snapshot in the person's vault. The new vault path does not write a server-side connection registry.
Older server-held connection rows and Plaid Items may remain. They are retired only after the per-environment cleanup procedure, migration, and disconnection checks succeed. Source review does not establish that a particular environment has completed those steps. The earlier Alpaca paper portfolio work is retained in the dated sprint history below; it should not be read as current live brokerage capability.
What iBrokerage is not
- Not a brokerage. Hussh does not custody assets. iBrokerage is the consumer surface; the user's accounts stay with their custodians.
- Not advisory. iBrokerage surfaces facts and savings; it does not make trades for the user without explicit per-action consent.
- Not connected to Fund A. Fund A is institutional/private; iBrokerage is consumer/public. They share a brand, not a regulator.
Sprint progression — May 2026 plan (historical)
This table preserves the original product sequence and target dates. Its states are not a current release or deployment report.
| Sprint | Goal | State at the May 2026 snapshot |
|---|---|---|
| Sprint 1 | Substrate — SQL migration, async Alpaca client, paper portfolio service, FastAPI routes, TypeScript frontend | Shipped 2026-05-05. Alpha Bets 27 model portfolio seeded. Nine-endpoint API. Platform-aware frontend service. |
| Sprint 2 | Account schema, user-owned data plane, consent protocol scaffolding | In progress. |
| Sprint 3 | The Aha Moment gate — Gmail connect → 60-second financial picture → dollar-denominated savings surfaced | Critical gate. Sprints 4-6 do not commence until Sprint 3 is real, fast, and beautiful. |
| Sprint 4 | Paper-trading simulator, Alpha Bets 27 follow/unfollow, achievement engine | Queued post-Sprint 3. |
| Sprint 5 | Live-trade architecture | Phase 2; deferred. |
| Sprint 6 | TestFlight beta launch | Target window: late July 2026. |
Private sprint artifacts and partner-facing product specs stay in the private artifact store until they are explicitly promoted. This public page is the canonical reader-facing synthesis.
Repo truth and north-star boundary
- Current source: Kai has a vault-based Plaid connection and refresh path for read-only financial context. A particular deployed environment requires its own route, migration, and Plaid Item cleanup evidence.
- Product direction: iBrokerage aims to prove One, Kai, and PCHP working together through a consented financial picture and a useful first number.
- Authority: no partner CRM, advisor system, or provider integration receives broad financial memory or a vault mirror without explicit consent, scoped export, and a retention policy.
- Evidence: the Aha Moment remains an acceptance target; this page does not claim that the one-minute flow or all legacy cleanup has passed UAT.
Sources
- One — relationship layer for the product direction.
- Kai — finance specialist driving the Aha Moment.
- PCHP — consent protocol behind import and provider reads.
- Aha Moment — onboarding acceptance target.
- Internal Hussh product notes, 2026-05-19 — original iBrokerage roadmap and acceptance target.
- Internal implementation review, 2026-09-23 — vault-based Plaid route and pending environment retirement evidence.