Support
Help
Version 2026-09-06 · Last updated
Short answers first, with links to the exact place in the product. If yours is not here, write to support@pythiaresearch.org — a person replies within two business days.
The product
How is Pythia different from Stock Rover, Simply Wall St, or Finchat?
Three things. First, open methodology: every score links to its formula and sources, not a black box. Second, two scoring tracks: the PAS, an academic 5-pillar score (sector × country z-scored), plus the PGS Guru Pyramid (18 guru signals tiered by conviction), so you get rigor and narrative side by side. Third, an AI analyst that cites the underlying filing for every claim, not vibes.
Can I trust the AI not to hallucinate?
Pythia's AI is grounded in retrieval: every answer is synthesized from filings, transcripts, and our metrics database, with citation chips you can click to open the source. If there isn't a source, the AI says so rather than inventing one. Nothing it writes is investment advice.
Where does the data come from?
Financials, prices, filings, and news come from Financial Modeling Prep, with every filing linked to the original document on sec.gov. Treasury yields and macro indicators come from FRED; FX reference rates from the ECB. Every row in our database carries its data source, filing date, and as-of timestamp, so you can always trace a number back.
How fresh is the data?
Prices refresh within 24 hours and are delayed or end-of-day, never real-time-for-trading; financials cycle through the full universe about every six days (with priority for recently-filed companies), and metrics recompute daily. Every metric you see carries its as-of timestamp, never silently stale.
How are the scores calculated?
PAS (Pythia Academic Score) combines five academic pillars (Profitability, Value, Quality, Momentum, Growth), each z-scored within its GICS sector × country, into a 0-100 composite. PGS (Pythia Guru Score) codifies guru signals from Buffett, Graham, Lynch, Greenblatt, Marks, Pabrai, Drucker, and Thorndike; the Guru Pyramid tiers companies into four bands from Deep Value to Overvalued. Click any score to see its inputs, formula, and source.
Can I export the data or use an API?
Every screen, watchlist, statement and company view exports to CSV and Excel, and you can download an archive of your own data from Settings at any time. The public API is paused until our data-vendor display licence covers redistribution; it will return for Enterprise plans first.
What is your refund policy?
7-day free trial on Premium, card required, cancel anytime. After that, a 14-day money-back guarantee on the first subscription payment. If Pythia is not pulling its weight in your process, email support@pythiaresearch.org for a full refund.
Your account
- Change email or password: Settings → Security. An email change is confirmed from the new inbox.
- Two-factor authentication and devices — the same page: enrol an authenticator, see every signed-in device, sign one out.
- Export my data: Settings → Profile, one JSON archive of everything we hold about you.
- Delete my account — the same page; it cancels any subscription, deletes your Stripe customer record and any brokerage connection, and removes your data.
- Cookies and analytics: Settings → Privacy; this is also the Do Not Sell or Share switch.
Billing
- Trial — 7 days on Premium, a card is required; cancel before it ends and nothing is charged.
- Invoices, payment method, cancellation: Settings → Billing opens the Stripe customer portal.
- Refunds — see the Terms; requests go to support@pythiaresearch.org.
- AI credits — bought in Settings → Billing; they never expire while the account exists and are not refundable.
Data
- Where a number comes from — every figure opens to its source and formula; the data sources page lists the vendors and freshness.
- Something looks wrong — use “Report a data issue” in the company toolbar; it sends us the ticker and the leaf.
- How scores work — the methodology page, and the Academy for the frameworks.
Reach us
Support: support@pythiaresearch.org · Security: security@pythiaresearch.org · Privacy: privacy@pythiaresearch.org · Everything else: contact page.