Portivo is a personal CIO for serious self-directed investors. You write your investment policy once. Every scan, every research memo, every thesis, every sell — checked against it, logged, and saved as your knowledge compounds.
Software for self-directed investors. Not investment advice, not personalized recommendations. You operate it; you decide.
This is not a robo-advisor and not a chat-with-a-friendly-AI app. It will feel like overkill if you don’t recognize yourself below — and like the thing you’ve been trying to build in spreadsheets if you do.
You already have the discipline. You know how to size, when to trim, how a thesis should read. The hard part is finding the hours — to read the transcript, triangulate it across three other things you saw, do the actual research, keep every thesis updated, and notice the trend before everyone is talking about it.
Three options exist today for someone managing serious capital. Each has a fatal gap.
Reconciling positions, watching drift, drafting theses in Notes, tracking which account holds what. Three hours a week of busywork that compounds in noise — not in insight, and not in any Knowledge Base you can search later.
Gap → no compounding memory$25K–$100K a year on a portfolio you understand better than they do. The relationship optimizes for their fee, not for the conviction you’ve actually built — and you still spend evenings on the work that matters.
Gap → outsourced judgment, kept workHelpful for a single question, useless across time. Chat tools answer in a vacuum: they don’t enforce sizing, don’t remember last quarter’s thesis, and don’t push back when you’re about to repeat the loss that hurt the most.
Gap → no policy, no continuityPortivo is a small set of commands that compose into a workflow. Each one writes to the same Knowledge Base — so research feeds theses, theses generate reviews, reviews become post-mortems, post-mortems sharpen the policy.
A two-session interview — life, money, behavior, history — produces your Investment Policy Statement. The CIO derives the rules; you don’t fill in numbers. Sizing caps, cash floor, bucket structure, forbidden actions, behavioral profile.
Seven slots, enforced for diversity: a new macro theme, a sector you don’t cover, two company stories, a non-US story, two wildcards. The brief connects dots — supply chain to commodity to inflation to consumer — not just the headlines you’d see anyway.
Drop a podcast transcript, a Substack post, an annual letter, a YouTube video, a Twitter thread, a research PDF. The CIO extracts the calls, agrees or pushes back with reasons, ties it to your existing positions and watchlist — then files it. Nothing dies in your tabs.
Fifteen to twenty-five tool calls, six search passes, multi-tier evidence, red-teamed by adversarial personas. Every research memo answers the same checklist: what is it, who owns it, who’s selling, what breaks the thesis, what would change your mind.
Secular or regime, with derived expiry. Hurdle rate vs. T-bills. Base / bull case grounded in milestones — not round numbers. ATR-based stops for regime trades. Fundamental invalidation for secular. Every tranche has a gate and a deadline.
Theses get reviewed quarterly, monthly, or weekly — the CIO sets the cadence by type. INTACT / EVOLVED / BROKEN. New versions are saved; old ones are preserved. You can always read how your thinking actually changed.
Outcome × process. A profitable trade with broken process is still a failure. A losing trade with disciplined process is still good investing. Lessons feed back into your Investment Policy Statement — your behavioral profile sharpens with every cycle.
The CIO calls real data — not vibes. Every claim is sourced and verifiable.
You don’t ask Portivo to “synthesize three articles.” You do your normal work — research a name, update a thesis, ask a question — and the CIO connects the dots from your Knowledge Base on its own.
Every transcript, memo, voice note, article, thesis, post-mortem — searchable by meaning (vector) and by structure (ticker, theme, date, source). Same content, two retrieval paths.
Run /research USD-JPY or just ask “start me on USD weakness.” The CIO checks your library first — before going to live data — and brings forward what you’ve already saved.
Druckenmiller said this in April. Marks framed it in March. Your own voice memo pre-empted both. You never told it to triangulate — it just does, because that’s what a CIO does when they know the book.
The structured commands are the spine. But you can bring the CIO anything — a half-formed thought, a thesis you want stress-tested, an urge you can’t shake. It will challenge you with your own rules, your own history, and the data.
Portivo is a real app — not a chat window pretending to be one. Every position, every thesis, every closed trade lives as a structured document with its full audit trail.
Positions across venues. FIFO realized PnL. IRR vs your custom benchmark.
Every time you break your own Investment Policy Statement — captured with severity. Monthly review. Patterns surface.
Bull / bear cases, sourced. Verdict + confidence. Event updates supersede.
v1 → v3 with diffs. INTACT / EVOLVED / BROKEN. History preserved.
FIFO lots. Behavioral tags. Post-mortems with outcome × process grades.
| Position | Bucket | Venue | % NAV | Cost | PnL | Thesis |
|---|---|---|---|---|---|---|
| NVDA | Core | Schwab | 7.2% | $112K | +62% | v3 INTACT |
| BTC | Tactical | Coinbase | 8.0% | $148K | +31% | EVOLVED |
| TAO | Spec | Coinbase | 3.4% | $78K | −12% | v1 INTACT |
| MUB | Core | Schwab | 11.2% | $268K | +4% | income |
| SF condo | RE | — | 28.4% | $680K | — | owner-occ |
| Date | Severity | Section | Rule | Detail |
|---|---|---|---|---|
| 04-29 | WARNING | Sizing | Single position cap | NVDA reached 8.4% intraday — auto-trim queued for next session |
| 04-22 | VIOLATION | Forbidden | Average down on broken thesis | Added to TAO after invalidation triggered. Override recorded; post-mortem required. |
| 04-14 | WARNING | Cash | Floor margin | Cash dropped to $81K — within $1K of $80K floor |
| 04-03 | OVERRIDE | Behavior | Cooling period | Manual override on cooling period for NVDA tranche 2 — reason logged |
Three personas attack the bull case before the verdict is locked. Each persona reads the full memo, then writes a critique. The CIO ranks which critique survives — and folds the strongest into the verdict-flipper.
“You’re calling this ‘decentralized AI compute’ but the unit economics look exactly like 2018 GPU tokens. Subnet emissions of 7,200 TAO/day at $340 = $2.45M/day in dilution. Compare that to your $20M ARR claim — that’s ~$60M / quarter in stock-based-comp equivalent against $5M in revenue. The ‘revenue’ is paid out in the same token used to buy compute. This is not a real economy; it’s a closed loop where price has to keep rising to keep the lights on. Your bull pillar #1 is mark-to-market, not cash.”
CIO note: this critique survived. Folded into the verdict-flipper as a 2-quarter emissions-vs-demand watch. If TAO/USD breaks $200 with this dynamic intact, thesis goes NEUTRAL regardless of price action.
“ETF pipeline at ‘6–9 months’ — same thing every alt-coin desk says. SEC has approved two non-BTC/ETH ETFs in 18 months. Your ‘regulatory tailwind’ is a base rate of ~12%, not the ~60% the bull case implies. Discount the catalyst by 4x.”
CIO note: valid — catalyst weight reduced from 25% to 8% in the verdict. Bull case re-scored from 5 strong pillars to 4 strong + 1 speculative.
“This is the same ‘decentralized [thing]’ narrative that cost you on FIL in 2022. You’re pattern-matching on a story you’ve already been wrong about once.”
CIO note: surface-level analogy. FIL’s revenue model was storage utilization without compute primacy; TAO’s subnets generate revenue from inference contracts. Different unit economics. Critique rejected, but the behavioral warning logged — your TAO sizing has crept past the 3% spec cap once already.
| Element | v2 | v3 | Status |
|---|---|---|---|
| View | — unchanged — | INTACT | |
| Catalyst (ETF) | Q3 2026 | Q4 2026 | DELAYED |
| T1 milestone | $50M ARR | $42M ARR | REBASED |
| Stop / invalidation | −65% catastrophic | −65% + 2-Q emissions | TIGHTENED |
| Capital justified | strong | weakening | FLAG |
| Position size | 4.5% | 3.4% | TRIMMED 25% |
| Asset | Held | PnL | % | Outcome × Process | Behavioral tag |
|---|---|---|---|---|---|
| SOL | 14 mo | +$41,820 | +41% | good × sloppy | can’t sell winners |
| UPST | 7 mo | −$22,400 | −28% | bad × disciplined | stopped at invalidation |
| ASML | 22 mo | +$54,210 | +38% | good × disciplined | trimmed at T1, T2 |
| RIVN | 3 mo | −$8,400 | −18% | bad × sloppy | FOMO entry |
Most AI tools treat your portfolio as data. Portivo treats your policy as law. Every suggestion is checked against the document on the left — and shipped with the receipts.
Edit it as a doc. Fork a template. Grow it month by month from your own post-mortems. The agent reads the same document, the same way, every morning.
Most “AI for investing” tools assume you’ll remember to use them. Portivo runs on rhythm — daily, weekly, monthly, quarterly, and the things that only happen when they happen. Each cadence has its own work and its own document trail.
“The discipline of an institutional CIO, applied to my portfolio, at the cost of a software subscription.”— Portivo, in one sentence
Portivo is software, not a financial adviser. It is a tool that self-directed investors use to organize their own research, write their own policy, and review their own decisions. It is not a registered investment adviser, broker-dealer, or financial planner.
Portivo does not custody your assets, place orders, or provide personalized investment advice. Nothing produced by the product — scans, research memos, theses, post-mortems, sparring transcripts — is a recommendation to buy, sell, or hold any security or asset. All output is informational only, generated from public data sources, and intended to support your own decision-making.
You are solely responsible for every investment decision you make. Consult a licensed adviser, accountant, or tax professional before acting on anything you read here. Past performance is not indicative of future results.
Join a small group of self-directed investors helping shape Portivo’s first public release.