Validated brief

The summary of the discovery cycle that made Felix a Go (April 2026), kept current as decisions land — last updated July 2026. The supporting research sits alongside: market landscape, competitive analysis, feasibility, assumption map, and the discovery decision log.

---

Executive Summary

Felix is an advisory-only wealth manager for self-directed investors: every account in one place, analyzed continuously by deterministic code, turned into a daily briefing and explained, structured actions — the user executes in their own accounts, and Felix verifies they did. The founding insight is that most of what a $20K/year advisor does is executing a known playbook, and nobody delivers that playbook as software that leaves the user in control. The original research framed this as a CSV-upload portfolio X-ray for US tech workers; the build has since grown into the full advisory loop, parameterized by jurisdiction (each user pinned to the US or India), currently deployed fee-free to friends and family while the core retention question gets tested. The product shape is in Surfaces; the system in the architecture; the story in Meet Felix.

---

Market Landscape

TAM/SAM/SOM

Confidence: Medium. TAM built from distributional tails — exact intersection of age + wealth band isn't directly published. SOM is speculative. All of it is US-only: the India market — now a first-class jurisdiction in the build — has not been sized.

Target Segments (Ranked)

1. Tech workers with equity comp (30-40% of TAM) — RSUs, ISOs, ESPP, concentrated stock. Highest pain, highest WTP, most reachable. Public company employees prioritized over startup.
2. Medical professionals (10-15%) — High income, student debt complexity, hard to reach at scale.
3. High-income dual earners (15-20%) — Large segment, moderate complexity, moderate pain.

Beachhead

Tech workers at public companies, age 28-34, $800K-$3M investable, RSU-heavy comp, based in CA/WA/NY/TX.

Validated as the strongest beachhead across every dimension: acute recurring pain (quarterly RSU vesting), high complexity (4-7 accounts, no tool handles it), high WTP ($200/month on subscriptions is normal), extremely reachable (Blind, Reddit, HN, company Slack), and intra-company network effects (everyone has the same comp structure). The beachhead is currently a go-to-market conclusion rather than a build input — the friends-and-family deployment is the first cohort, and equity-comp features remain deferred.

Market Timing

Strong tailwinds: self-directed investing surging among under-40s, robo-advisors proved the category but left a gap, AI makes advisor-grade software feasible, $84T generational wealth transfer ahead, rising tax complexity for high earners.

Key headwind: trust deficit for AI financial advice (25% of Americans would trust it, but skews much higher for target demo).

---

Competitive Landscape

Positioning Map

                    YOU DECIDE (Advisory Only)
                           ^
                           |
     ProjectionLab         |        *** WHITE SPACE ***
     Boldin                |        [Felix]
                           |
     Kubera                |        Mezzi (moving up-right)
     Monarch               |        Astor · PortfolioPilot
     Origin (broad/shallow)|        Empower (free tools)
                           |
  Frec (TLH only)         |
  Secfi (equity only)     |
                           |
SINGLE -------------------|----------------------- FULL ADVISOR
FEATURE                   |                        REPLACEMENT
                           |
     Schwab Intelligent    |        Betterment
                           |        Wealthfront
                           |        Arta (accredited only)
                           |        Vanguard Digital + Personal
                           |        Empower (managed)
                           |        Human Advisors
                           |
                    THEY MANAGE YOUR MONEY

The upper-right quadrant is contested but unowned. No product yet combines comprehensive advisory (portfolio analysis + TLH + asset location + rebalancing + equity comp + retirement) with an advisory-only model where the user keeps control. It is no longer empty, though: within a quarter of the original research, Mezzi repriced upmarket and registered as an SEC fiduciary, and Astor launched at $15/month.

Key Competitors

Pricing Gap

Massive gap between $100-$200/year (DIY planning tools) and $2,500-$10,000/year (managed robo or advisor). The $200-$2,500/year range for ongoing, personalized, comprehensive advisory was nearly unoccupied when the research began — Mezzi's repricing into it is the first occupancy, and confirmation that this is where the category prices.

Defensibility

Strongest moat: incumbents' business model conflict. Wealthfront, Betterment, Empower, and Arta all monetize via AUM. Building a comprehensive advisory-only product would cannibalize their core revenue. They won't do it.

Secondary moats: Comprehensiveness (full playbook with edge cases is 12-18 months to replicate), switching costs (financial history builds over time), execution speed.

Honest assessment: Moderate defensibility. The business model conflict is real and durable. The technical moat is not deep — a well-funded team could replicate features. The competitive risk is from other new entrants, and Mezzi and Astor proved in one quarter how fast the space moves. Felix's counter is depth and trust: deterministic analytics, a daily loop, and verified actions rather than a chat layer over aggregated data.

---

Top Risks — and Where Each Stands

The full map — 22 assumptions across 6 categories — is in the assumption map.

---

Feasibility Verdict

How the MVP resolved

The original research weighed two MVP shapes — a scorecard with top-3 recommendations, or a portfolio X-ray with TLH flagging. Neither was built as scoped: the retention analysis (a one-time X-ray doesn't sustain a subscription) pushed the design straight to the full advisory loop — nightly refresh, deterministic fact catalog, daily AI briefing with structured actions, a verification loop that matches accepted actions against real transactions, news, and chat. The X-ray analytics survive inside the analytics engine. Data ingestion resolved the same way the feasibility work recommended: CSV first, brokerage-native sync layered on, Plaid deferred.

---

The Bull Case

Why this should be built:

1. The white space is real and verified. No product delivers the full advisor playbook in an advisory-only, user-controlled format — and the first entrants repricing into the gap confirms the category rather than closing it.

2. The beachhead is ideal. Tech workers with equity comp have acute, recurring pain (quarterly RSU vesting), high complexity (4-7 accounts, no tool handles it), proven willingness to pay for software, and cluster in reachable communities. Intra-company network effects could drive organic growth.

3. Incumbents can't build this. Wealthfront, Betterment, Empower, and Arta all monetize via AUM. An advisory-only product that says "keep your money at Schwab" cannibalizes their revenue. This business model conflict is a real, durable moat.

4. The timing is right. AI makes advisor-grade explanations feasible. Self-directed investing is surging. Tax complexity for high earners is rising. Fee awareness is at an all-time high. $84T in generational wealth transfer is ahead.

5. The economics are attractive. The expensive AI work runs once and fans out; a marginal user costs roughly $20/month all-in. At subscription pricing the business is profitable per user from day one, with an expansion path toward the full playbook at $50-$150/month.

6. The core insight is sound. Most of what a wealth manager does is executing known best practices consistently. That's automatable. The target demo already suspects this. The fee arbitrage narrative ($20K/year advisor → $200-$360/year software) is immediately legible.

---

The Bear Case

Why this should NOT be built:

1. Retention is the existential question and it's still unproven. The comparable data point is grim: Empower's free dashboard converts ~2% to paid; the other 98% looked at their portfolio, said "interesting," and stopped engaging. The daily loop is designed specifically against this — but designed is not proven, and the friends-and-family cohort has to show people actually come back every morning.

2. The advice-to-action gap may cap the market. Advisory-only products consistently show lower engagement than advisory + execution products. "Here's what's wrong" without "click here to fix it" may feel like homework, not relief. The verification loop reframes this — actions close with evidence, nudge, or expire — but the structural pull toward execution (and the AUM business that comes with it) never fully goes away.

3. The quadrant is filling. Mezzi repriced upmarket and registered as an SEC fiduciary within a quarter; Astor launched cheap and conversational; general AI assistants keep raising the free floor. First-mover advantage in the empty quadrant is gone — the race is now on depth and trust, and if Wealthfront ever deepens Path into full advisory, the window narrows further.

4. Subscription pricing may not support a venture-scale business. At $20/month average and 25,000 users after 3 years, that's $6M ARR. Meaningful, but not venture-scale without higher pricing for the full playbook, a larger TAM (India is the untested expansion), or execution/AUM — which would change the product's nature. Mezzi's $299-$1,499/year tiers suggest the ceiling is higher than first modeled, but it's their experiment for now.

5. Regulatory risk is real but currently quiet. The SEC withdrew its proposed predictive-data-analytics rule in June 2025; AI-driven advice is governed by existing Advisers Act fiduciary obligations, with enforcement focused on "AI-washing." A future AI-specific framework would be a heavier lift than RIA registration — a watch item, not a blocker.

6. The target demo is the hardest to convert. High-earning, financially literate 25-35 year olds are the most likely to think "I could just do this myself with a spreadsheet." They're also the most likely to comparison-shop, read every review, and churn if the product doesn't deliver continuous, surprising value. They're ideal users if the product is great — and the first to leave if it's merely good.

---

Open Questions

1. Does the daily loop actually retain? The design answer exists — nightly re-pricing, the briefing, actions, verification. The friends-and-family deployment is the experiment that answers it with behavior instead of theory. This remains the most important question in the project.

2. What does monetization look like when it comes? Pricing is deferred, the RIA question (and India's SEBI equivalent) is parked with it, and the market has moved since the research — Mezzi's $299-$1,499/year band is one data point that the ceiling is higher than the original $10-$30/month framing. A one-time-report model remains the honest fallback if retention disappoints.

3. Is India a market or just a jurisdiction? The build supports it; the research doesn't cover it. Sizing, competitive landscape, and SEBI's advisory regime are all open work before Indian users are more than friends and family.

4. When does the beachhead come back? The RSU tech-worker wedge is validated but dormant — no equity-comp features exist yet. Whether it's still the opening move gets decided when Felix turns to growth.

---

Recommendation

Go — and the Go has held.

The opportunity is real: genuine white space (now contested, still unowned), a sharp beachhead in reserve, incumbents structurally unable to compete, and favorable timing. The bear case is serious — retention and the advice-to-action gap are legitimate risks — but they're testable risks, and the current track tests them at almost no cost: a solo, fee-free deployment running at roughly $85–95/month, sequenced so the one existential question — do people come back every morning? — gets answered with real behavior before any scaling decision.

What matters next, in order: prove the daily loop retains; do the India research the original cycle never did; then make the monetization call — pricing, RIA registration, and whether the RSU wedge is still the way in.