How Reprisa thinks — and how to build on it.

Reprisa reasons over 120+ years of market history — prices, rates, credit, valuation, breadth, sentiment, and volatility across asset classes — combined with live quotes pulled at the moment you run a report. It scores today's conditions against that history on a fixed 15-factor structural checklist. It does not use your personal financial accounts or trading activity as an input.

Every report is temporally grounded. It pulls today's live quotes and the current date, and the engine anchors every forward-looking reference — next earnings, Fed meetings, catalysts — relative to today rather than to something half-remembered from training. Where it has no live data for a specific point, it says so plainly instead of inventing specifics.

No. Reprisa is an informational research tool. Its reports, parallels, and signals are historical and structural context to inform your own thinking — not a recommendation to buy, sell, or hold anything. Past performance is not a guarantee of future results. See the Terms for the full disclosure.

You can, and for a quick take you probably should. The difference is discipline, and it's checkable. Reprisa reports are assembled from a locked fact table before a single sentence of prose is written — the numbers can't drift mid-report. Live quotes are pulled and timestamped at the moment of generation, so you know exactly what the analysis saw. Where the engine has no data for something, the report says so instead of papering over the gap. And every historical comparison is scored on the same fixed 15-factor checklist, which makes results reproducible and comparable across reports — run the same question twice and you can audit why the answers agree or differ. A general chat assistant gives you none of those guarantees. The changelog tracks these mechanisms as they ship, and the method section above explains the scoring.

You can cancel any time from your account settings — cancellation takes effect at the end of the current billing period, and you keep access to paid features until then. Subscriptions renew automatically until cancelled. Except where required by law, payments are non-refundable for partial billing periods, so the practical move is to cancel before renewal rather than after. The free plan needs no card at all, so you can evaluate the product fully before paying. Full terms in the Terms.

Three kinds of sources. Live market data — quotes, index history, and market news — comes from established public market-data feeds, pulled at generation time and timestamped in the report. Company filings and earnings figures come directly from SEC EDGAR, quoted verbatim from the filings. Charts on ticker pages are embedded TradingView widgets, with TradingView's own attribution shown in the widget. The historical record the engine reasons over — prices, rates, credit, valuation, breadth, sentiment, volatility across 120+ years — is the engine's research corpus, scored against today on the fixed 15-factor checklist. We don't use your personal financial accounts or trading activity as an input.

The 1–10 impact score you see on news headlines and Daily Recap items is Echo's judgment of how much an item matters to the broad US market right now — not a quantitative market metric. Several times a day the engine reads the incoming headlines alongside current broad-market moves and assigns each a score from 1 (minor) to 10 (major, market-moving), with a neutral one-line "why it matters." The ranking runs under binding no-advice rules: it never predicts direction, never recommends action — it only judges relevance. Scores are model judgment applied consistently on one scale across the product, so an 8 on the news feed means the same thing as an 8 on a recap item. When the ranking engine is unavailable, headlines fall back to an unranked neutral 5 rather than an invented score.

Echo is Reprisa's research engine — the prompting, the 15-factor structural scoring, the temporal grounding, and the orchestration that turns a question into a sourced answer. It runs at three depth tiers: Echo Flash for instant single-pass reads, Echo Core for adaptive multi-factor reasoning, and Echo Max for maximum-depth, multi-step reasoning. Your plan sets how deep it runs. Echo is a capability, not a claim that we trained our own foundation model.

Click any stock — in a report, your watchlist, or the news feed — and a full terminal opens: Echo's plain-language read of the name, the fundamentals that matter, an interactive chart, the latest news, and any recent signals, on one screen. The 0–100 bullish meter is Echo's read of the fundamentals, not a target and not a recommendation. It is context to inform your own judgment — not financial advice.

Deep Analysis is a longer-form report. Instead of a quick parallel, Echo works in the background to write a complete, structured research document — an overall rating, a rating and key points for every section, and the evidence behind each. You kick it off and come back to a finished brief, saved to your history and exportable. The ratings judge the strength of the analysis and the evidence, never a price or an outcome, and nothing in the document is a recommendation to act.

A live market bar sits above a news feed that Echo ranks by market impact, scoring each headline and writing a neutral one-line note on why it matters — what the news concerns and which part of the market it touches, never what to do about it. Click any story for an instant brief. The same ranked snapshot is what grounds a fresh report in today's tape. Impact scores are a read of relevance, not a directional call.

Yes. Reprisa runs pattern alerts, unusual-activity flags, insider-filing alerts, and upcoming-earnings alerts on the tickers you follow, plus a pre-market Morning Brief — delivered in-app and by email, with per-item thresholds you control. Alerts are historical and structural context, not recommendations, and past performance is no guarantee of future results.

Every report is saved to your history and exportable as a polished PDF. You can also create a read-only share link so someone can view a report without an account. Shared reports are visible to anyone with the link; nothing else in your account is exposed. The Privacy Policy covers how shared and stored data is handled.

The two setups scanned daily — a pullback to a rising 20-day moving average and an extension-fade at a declining 200-day moving average — have rules that were backtested across a decade of varied market regimes. Every alert carries the setup's transparent base rates so you see the historical context, not a promise. Base rates are averages, not predictions, and past performance is no guarantee of future results.

Yes. You can cancel from your billing settings whenever you like; your plan stays active through the end of the current billing period and then reverts to Free. No cancellation fees, no lock-in.

No. All payments run through Stripe, and card details are handled and stored by Stripe — never on Reprisa's servers. See the Privacy Policy for how billing data is processed.

Your reports, watchlists, and workspaces are scoped to your account — private unless you choose to post something to the community feed. Nothing you research is shared with other users by default. The Privacy Policy covers data handling and subprocessors in full.

You can, and for general questions a chatbot is useful. A research report is a different job. Reprisa pulls live quotes and today's impact-ranked headlines at the moment of generation, so the analysis is anchored to the current tape rather than to wherever a model's training data ends. Every parallel is scored on the same fixed 15-factor checklist and the score is shown — 6 of 15 reads as exactly that — so you can audit why a match was called a match. Where there's no live data for a specific point, the report discloses the gap instead of inventing specifics. And the output is a structured, saved, exportable report you can re-open and defend later — not a chat transcript.

Subscriptions renew automatically each billing period until cancelled. You can cancel anytime from your billing settings; cancellation takes effect at the end of the current period, and you keep access to paid features until then. Except where the law requires otherwise, payments are non-refundable for partial billing periods — cancelling stops the next charge rather than refunding the current one. The full billing terms are in the Terms.

Live quotes and historical price series come from public market-data providers — quotes are pulled fresh when a report runs, and the same providers feed the charts and the backtested signal scans. US macroeconomic series — rates, inflation, recession indicators — come from public federal data (FRED). Headlines come from a live market-news feed that Echo ranks by market impact before it grounds a report. There's no proprietary exchange feed and we don't claim one: the edge is the method — the 15-factor scoring and the temporal grounding — not exclusive data.

The 15 factors are structural rather than surface-level — monetary policy direction, asset valuation level, market breadth, credit conditions, the volatility regime — so a 1970s parallel and a 2020s situation are compared on the same dimensions even though market structure changed in between. Regime mismatch is also exactly where the honest score shows up: a period that rhymes on sentiment but sits in a different rate and liquidity regime simply matches fewer of the 15 factors, and the report says so rather than dressing the parallel up. The method section lists all 15 factors.