Model Score
Represents the strength and quality of systematic model evidence in the validated research dataset. Higher values indicate stronger model evidence within that dimension; they are not an instruction to trade.
METHODOLOGY TRANSPARENCY
Understanding AI INVEST scores
The public website shows evidence scores and confidence metrics. It does not translate those values into final Buy, Sell, Strong Buy, Strong Sell or Neutral verdicts.
Technical Score
Summarizes technical evidence such as trend, momentum, volatility and market structure. It is shown independently so technical conditions can agree or disagree with other evidence layers.
Fundamental Score
Represents evidence on business quality, growth, profitability, balance-sheet strength, cash quality, valuation and related company fundamentals, subject to available validated data.
Composite Score
A normalized integrated research metric from the validated framework. Its purpose is comparison and change tracking. It is not mapped publicly to an investment verdict.
Data Confidence
Measures evidence completeness and data quality. Missing or stale evidence can reduce this score. A high research score with low data confidence should be interpreted differently from the same score with strong data confidence.
AJ Confidence
Represents confidence in the overall research evidence state. It is separate from the direction or action a user may independently choose.
How to use the public scores
1Compare dimensions. Look for agreement and disagreement between model, technical and fundamental evidence.
2Check confidence. Data Confidence and AJ Confidence help judge how much validated evidence supports the scorecard.
3Review history. A score's direction and magnitude of change can be more informative than one isolated reading.
4Decide independently. AI INVEST does not publish the final action label on the public website. Users determine what action, if any, is appropriate for them.
Missing-data and audit controls
AI INVEST is designed to neutralize or flag unavailable evidence rather than fabricate it. The daily public dataset is generated from an Audit Excel workbook, validated server-side, staged for review and only published after reviewer approval. Look-ahead controls, duplicate checks, stale-date guards, score-range controls and universe-size checks are among the publication controls.
Methodology details shown publicly are intentionally explanatory rather than formula-complete. Proprietary factor weights, internal directional labels and other confidential implementation details are not published on this pre-launch website.
AI and conflicts disclosures
AI INVEST uses automated and AI-assisted analytical processes as part of its research infrastructure. Outputs remain subject to data-quality, validation, governance and publication controls.
AI INVEST (through Jamuda Technologies LLP), its stakeholders, partners, directors, employees, affiliates or related persons may or may not hold, trade or otherwise have an economic interest in securities mentioned on the site. Such interests may change over time. Required security-specific disclosures remain separately applicable where required.
No verdict mapping.AI INVEST does not define a public rule such as “80–100 = Strong Buy.” Numeric scores are research evidence, not a public final verdict.