Intratio Tutorial & Complete Feature Guide

1. Onboarding & Account Basics #

1.1 Create Account & Verify Email #

Sign up using email or supported social providers (Google / Microsoft / Facebook). Email verification is mandatory for API access and quota tracking.

1.2 Subscription Tiers #

You need the Optimize entitlement for all programmatic endpoints and advanced analytics modules. Visit Pricing to upgrade.

1.3 Retrieving Your API Key

Every request must include an X-API-Key header. Base URL for this session: https://www.aistockfinder.com/api/v1.

1.4 Core Data Objects

• Company: Fundamental snapshot + trailing numeric metrics.
• Signal: Forecast value per horizon (3_months, 1_month, 2_weeks, 1_week, 3_days) and optionally alpha-adjusted variants.
• Portfolio Optimization: Mean-variance style solver leveraging forecast-derived expected returns, constraints and objective variants.
• Analytics Metrics: Distributions, correlations, outliers, trends, change detection, percentiles and peers aggregation.

2. Navigating the User Interface #

2.1 Stock Forecast Pages

Each ticker page consolidates latest multi-horizon forecasts, historical signal trajectory, fundamental highlights and percentile placements versus sector & industry.

2.2 Screeners

Forecast Screener: Rank by direction & intensity across horizons. Company Screener: Filter metrics using range syntax (e.g. Gross_Margin_percentage__gte=40).

2.3 Following Tickers

Use follow/unfollow buttons or endpoints to curate a watchlist. Improves signal change triage on login.

3. Core Forecast & Metric Concepts #

3.1 Forecast Horizons

Horizon keys: 3_months, 1_month, 2_weeks, 1_week, 3_days. Alpha versions normalize relative to baseline risk/benchmark behaviour.

3.2 Signal Quality & Change

Change compares successive dates; Trend accumulates over a window (fast SQL variant). Use both to detect emerging acceleration.

3.3 Percentiles

Percentile = percentage of observations strictly below the company value. Provided for All, Sector and Industry contexts.

3.4 Aggregations

Sector & industry endpoints return primarily Median to reduce outlier distortion. Dispersion endpoints show spread (std, IQR) enabling factor crowding detection.

4. API Quick Start #

Base URL: https://www.aistockfinder.com/api/v1 — Add trailing slash & query parameters as needed.

curlcurl -s -H "X-API-Key: YOUR_API_KEY" https://www.aistockfinder.com/api/v1/signals/top/?horizon=3_months | jq

pythonimport requests
API_KEY='YOUR_API_KEY'; BASE='https://www.aistockfinder.com/api/v1'
r = requests.get(f"{BASE}/companies/AAPL/", headers={'X-API-Key':API_KEY})
print(r.json()['ticker'], 'loaded')

jsconst API_KEY='YOUR_API_KEY'; const BASE='https://www.aistockfinder.com/api/v1';
fetch(`${BASE}/metrics/distribution/?metric=Gross_Margin_percentage`,{headers:{'X-API-Key':API_KEY}})
  .then(r=>r.json()).then(console.log);

4.1 Essential Endpoints

4.2 Authentication Pattern

Use a thin wrapper to attach headers; rotate keys by regenerating in profile (coming soon) without downtime.

5. Core Analytical Workflows #

6. Portfolio Optimization #

The optimizer maps selected tickers to expected returns derived from the latest horizon forecast (raw or alpha). Objectives:

• MaximizeReturnToRisk (maximum Sharpe)
• MinimizeRisk (efficient risk given target return)
• MaximizeReturn (efficient return given target risk)

Optional constraints: weight bounds, total cash out (capital deployment), target portfolio beta.

pythonimport requests, json
payload={
  "tickers":["AAPL","MSFT","GOOGL"],
  "horizon":"3_months",
  "alpha":false,
  "objective":"MaximizeReturnToRisk",
  "weight_bounds":[0,1],
  "target_risk":10,
  "target_return":10
}
res=requests.post("https://www.aistockfinder.com/api/v1/portfolio/optimize/",headers={'X-API-Key':'YOUR_API_KEY'},json=payload)
print(res.json()['performance'])

6.1 Interpreting Output

6.2 Best Practices

• Exclude tickers missing forecasts (endpoint returns errors for missing).
• Apply metric filters for quality (e.g. Debt_per_Equity upper bound).
• Rebalance when /signals/change/ indicates drift beyond threshold.

7. Advanced Analytics Modules #

The Analytics Dashboard renders interactive Plotly visuals powered by the same endpoints:

• Sectors & Industries Ranking: Median aggregation by selected metric.
• Signals Distribution: Central tendency & tail profile for horizon.
• Metric Distribution & Outliers: Robust trimming & hist fences.
• Correlation Matrix: Multi-metric relationship map.
• Scatter & Regression: Quick factor scatter with linear fit & correlation.
• Percentile Compare: Multi-ticker relative positioning across metrics.
• Signals Change & Trend: Short term shock vs multi-bar momentum.
• Company Snapshot: 20-key metrics vs peer median.

7.1 Extending Analytics

Copy network calls from browser DevTools (all /api/v1/...)—compose bespoke research notebooks using identical JSON contracts.

8. Monitoring, Watchlists & Risk Flags #

Design a lightweight monitoring loop:

pythonimport requests, time
API='YOUR_API_KEY'; BASE='https://www.aistockfinder.com/api/v1'
WATCH=['AAPL','MSFT','NVDA']
while True:
  ch=requests.get(f"{BASE}/signals/change/?horizon=3_months",headers={'X-API-Key':API}).json()['results']
  flagged=[r for r in ch if r['ticker'] in WATCH and abs(r['change'])>0.05]
  if flagged: print('Rebalance candidates:', flagged)
  time.sleep(3600)

Combine with /metrics/outliers/ to dynamically exclude deteriorating quality names.

9. Data Integrity & Limitations #

• Daily forecast refresh—intraday calls may not shift values until next batch.
• Median aggregation over raw averages to damp outlier skew.
• Percentiles strict-below methodology (ties cluster at same rank threshold).
• Missing metrics return null; propagate carefully in custom analytics.

10. Roadmap & Extensibility #

11. FAQ #

How often should I re-run optimization?

Common cadence: daily after forecast refresh, plus on significant /signals/change/ divergences.

Raw vs Alpha forecast?

Alpha attempts to isolate stock-specific return component; use in diversified baskets to reduce systematic drift.

Why medians everywhere?

Robust central tendency—less sensitivity to single distortive outlier values, improving comparative stability.

Percentile vs Rank?

Percentile gives continuous 0–100 scale; you can derive rank by ordering on value if discrete ordering required.