The MosaicIQ Platform

Technology that augments judgment

Proprietary, patent-pending, AI-native research software, built and maintained in-house. The platform structures our coverage universe according to a diverse, custom-built sector taxonomy tailored to the fund’s research framework. Every analytical output is timestamped and preserved — each decision reconcilable against the information available at the moment it was made.

End-to-end

One platform, three core workflows

From identifying a candidate, to sizing the position, to grading the decision after the fact.

1. Research & selection

Every covered company gets a living thesis — bull, base, and bear scenarios with testable catalysts and risks. Valuation adapts per name, a dual conviction read disciplines the universe, and a thematic layer maps hypotheses to beneficiaries and constrained names.

2. Entry, sizing & construction

Daily signals flag entry readiness, timed by a technical overlay that never overrides fundamentals. Exposure analytics catch hidden concentration, factor, and theme risk. Sizing follows conviction-weighted recommendations and a decision scorecard for each add, trim, or exit.

3. Performance measurement

Because every thesis, score, and decision is timestamped, the platform grades itself: theses are measured against outcomes, automated backtests recalibrate scoring weights toward what predicts returns, and a behavioral-bias layer audits the fund’s own decisions.

What makes it different

Discipline, compounded

The platform’s value is the discipline it enforces and the longitudinal record it builds over time.

Living investment thesis

Theses are versioned data models, not static documents; catalysts and risks carry status, evidence weight, and a full transition history, so a thesis evolves with the facts rather than going stale.

Evidence-to-estimate bridge

When new research shifts a tracked catalyst or risk, the platform derives a calibrated, audit-trailed estimate adjustment — connecting qualitative evidence to quantitative forecasts without losing the record.

Closed validation loop

Price targets, scenario probabilities, and implied upside are measured against what actually happened, and the lessons feed back into the scoring weights — the model is built to learn from its own mistakes.

Behavioral-bias audit

The fund’s own conviction, estimate, and position trajectories are checked against the CFA Institute bias framework, surfacing anchoring, overconfidence, and herd behavior in its own decisions.

On the roadmap

Toward multi-agent reasoning

We are extending the platform toward coordinated reasoning agents — a researcher that gathers evidence, an analyst that forms a view, and a synthesis step that reconciles them — to surface cross-sector hypotheses a single analyst might take weeks to connect.
Discuss the approach →

The platform does not make decisions. It automates the observation, measurement, and bookkeeping no single analyst can sustain across the full coverage universe — freeing capacity for the high-judgment work that drives returns.

For informational purposes only. Nothing on this site is an offer to sell or a solicitation of an offer to buy any security. See Disclosures.