Convert every sustainability disclosure into a governed, evidence-linked, approval-tracked record. From raw marketing text to supervisory-ready audit pack — with AI doing the extraction and humans staying in control.
5-step
End-to-end workflow
CT/ET
Standardised taxonomy
0–100
Deterministic risk score
1-click
Audit pack export
Every sustainability claim is a structured object, not a sentence. The AI reads your fund documents — factsheets, websites, KIDs, SFDR templates — extracts claims, and classifies each one against a standardised taxonomy.
Exclusion Constraint
Fossil Fuels
Net-Zero Alignment
Paris 1.5°C
Stewardship
Engagement
ESG Integration
Score Threshold
Each claim type has a defined evidence rule matrix. Once a claim is classified, the system auto-generates the specific evidence checklist required to support it — no manual mapping, no missing items.
Screening methodology doc
OKThreshold definition
OKHoldings verification Q4
ExpiringData vendor confirmation
MissingAssurance statement 2025
OKEvery claim receives a risk score from 0–100. Scores are deterministic and explainable — driven by evidence completeness, claim specificity, and flag triggers. No black box.
CLAIM RISK SCORE
42
Human approval is not optional — it is structurally required. Every claim passes through a named approver via Microsoft Teams with SLA tracking, version control, and an immutable decision record.
Claim extracted from factsheet v3.2
Risk score 42 — HIGH RISK flagged
Evidence checklist submitted
Compliance review approved
Audit pack generated and locked
The Trust & Risk Engine is a Power Platform–native application with a Next.js AI endpoint. No exotic infrastructure — deployable into your existing Microsoft 365 tenant.
Claim Intake Form
URL / upload / paste
Risk Dashboard
Fund-level overview
Evidence Repository
Status & expiry
Approver Queue
SLA + decision trail
AI Trigger Flow
Calls Next.js endpoint
Evidence Checker
Auto-populates checklist
Approval Dispatcher
Teams task + SLA timer
Audit Pack Generator
PDF export on approval
Claim Extractor
LLM extraction + CT/ET codes
Evidence Mapper
Rule matrix lookup
Risk Scorer
Deterministic 0–100
Flag Explainer
Plain-language reasoning
ClaimRecord
Versioned claim entity
EvidenceItem
Linked doc + expiry
ApprovalDecision
Immutable decision log
TaxonomyConfig
CT/ET rule matrix
AI assists, humans decide — every approval is a named person.
Every action timestamped and locked in Dataverse audit tables.
Every risk flag has a plain-language reason. No black boxes.
Built on Microsoft Power Platform — your IT team already knows it.
No new vendor approvals for exotic tools. Dollarbaz is built on Microsoft Power Platform — infrastructure most enterprise compliance teams already operate inside.
Approval tasks are delivered directly into Teams, where compliance teams already work. No context-switching — reviewers see the full claim, risk score, and evidence in one adaptive card.
Operational dashboards for compliance managers and fund governance teams. Track evidence completeness, review SLAs, and flag concentration risk across your fund range in real time.
All claim records, evidence items, approval decisions, and taxonomy configurations live in Dataverse — your Microsoft 365 tenant. You own your data, and nothing leaves your environment.
The AI scoring engine is exposed as a documented REST API. Embed claim extraction and risk scoring into your own portals, fund management systems, or distribution platforms.
All data stays within your Microsoft 365 tenant. Dollarbaz does not store or process your fund data on external servers.
Data Sovereignty ✓A concierge pilot configured for your fund range. Fixed fee. Measurable KPIs. Evidence completeness, review cycle time, and a supervisory-ready audit pack on day 30.
Not legal advice · AI-assisted, human-controlled · Data stays in your tenant
