Oxeye features · a product of Oxeye Technologies LLC
Everything in Oxeye earns its place on the screen.
Oxeye is the flagship iOS product built by Oxeye Technologies LLC. We ship a small, opinionated set of tools — each of them traces back to a concrete answer you'd actually want from the data. Here's the tour.
Recommendations
ML-ranked "Strong Buy / Buy" shortlist, refreshed every trading day
The Home tab shows only stocks that cleared our quality filters (market cap ≥ $100M, P/E between 0 and 500, debt/equity ≤ 5, ROE non-negative, liquidity ≥ 250K average daily volume, at least 2 years of trading history) and scored in the top two of five quality tiers.
Each card shows the ticker, company name, the rating, the predicted 90-day forecast, and the backtest-derived confidence. No news feed. No infinite scroll. Just the list.
- Five-tier rating system: Strong Buy, Buy, Neutral, Sell, Strong Sell
- Filterable by sector and rating on the Search tab
- Sortable by ML score, forecast change, market cap, or dividend yield
Market Insight — Money Flow
Where is capital actually moving this week?
The Insights tab tracks weekly AUM changes across every US-listed ETF (more than 4,600 of them), weighted by each ETF's sector and country exposure. The result is a clean line chart of net dollar flow into and out of sectors and countries — the aggregate signal under the noise of headline performance.
- Toggle between sector-level and country-level views
- Tap any data point to drill into that week's inflow / outflow bar chart
- Tap any bar to see the specific ETFs and sub-exposures driving it
- Sector heatmap shows weeks × sectors with color-coded flow intensity
- Pipeline-generated AI summary calls out the week's highlights and sentiment
Stock detail
Forecasts, fundamentals, and an LLM outlook on one screen
Tap a ticker and you get the whole picture: a 90-day ARIMA forecast with upper and lower confidence bounds charted against the last three months of actual prices, the fundamentals that drove the composite score, and an AI outlook that reads the model's output in the context of today's macro.
- Forecast chart with tap-to-inspect tooltips on every data point
- Composite score breakdown: which features helped, which hurt
- LLM-generated bull case, bear case, and key catalysts
- Confidence level adjusted for VIX and yield-curve state
- Back-tested accuracy for this specific ticker over the last 2 years
Also in the app
The rest of the toolkit
Portfolio metrics
Add tickers to your watchlist and Oxeye computes a historical Sharpe ratio, forward Sharpe (using our ML forecasts), annualized volatility, and pairwise correlation — all benchmarked against the live 3-month Treasury rate.
Watchlist with price tracking
We record the price at the moment you add a ticker and show gain/loss since that entry, alongside the ML forecast, so you always know whether a call has worked.
90-day reminders
When you add a stock to your watchlist, Oxeye schedules a local notification 90 days out so you can compare the forecast to the actual result and decide whether to keep trusting the signal.
"AI called it" celebrations
When a forecast hits its target on time, Oxeye flags it — and when one misses, it flags that too. The full accuracy log is visible per ticker so the score can't cherry-pick its wins.
Shareable prediction cards
Generate a clean, branded image of any stock's rating, forecast, and confidence level to share with friends or your own accountability group. No tracking pixels, no link-shorteners.
Sign in with Apple or Google
Firebase Authentication with Apple and Google. We store no passwords, and on iOS we enforce Firebase App Check so only the real Oxeye app can call our API.
Sub-100ms reads
The app talks exclusively to a Firestore-backed API that serves pre-computed results. No spinners, no stale dashboards, no "please wait while we query BigQuery".
Native dark mode
Designed to Apple Human Interface Guidelines using iOS system colors and SF-style typography. Pure-black background means true power savings on OLED devices.
Offline-friendly
Your watchlist and cached recommendations are available without a connection. We don't ask permission for anything we don't actually need to do the job.
Quality filters
What has to be true before a stock even gets scored
Before a ticker reaches the ranking, it has to clear a set of hard filters. These rules are chosen to keep penny stocks, zombie companies, and illiquid names out of the recommendation set.
| Filter | Threshold | Why |
|---|---|---|
| Sector | Not null | Required for sector-normalized P/E |
| P/E ratio | 0 < P/E ≤ 500 | Profitable, not extreme valuation |
| Market cap | ≥ $100M | No penny stocks |
| Debt / equity | ≤ 5 | Not overleveraged |
| Return on equity | ≥ 0% | Not destroying shareholder value |
| Current ratio | ≥ 0.5 | Adequate short-term liquidity |
| Price / book | > 0 | Positive book value |
| Trading history | ≥ 504 data points | At least two years of daily data |
| Average volume | ≥ 250,000 shares / day | Tradable at reasonable cost |
Want to see it running?
TestFlight access is available on request.
The Oxeye iOS app is currently in closed beta. If you'd like to be added to the next TestFlight group, drop us a line.