Regulatory outputs
Structured submissions such as XML, plus supporting packs and evidence.
Platform
One controlled workflow from source data to final report.
Datox connects to the data you already have, automates the extraction, enrichment, mapping, calculation and validation work between systems and spreadsheets, and produces reviewed, traceable reporting output, regulatory, financial or operational.
The Datox workflow
Connect. Extract. Enrich. Transform. Validate. Collaborate. Deliver.
The same seven stages run behind every Datox reporting workflow. Agents operate across the stages; governance, lineage and access control sit underneath all of them.
Connect
files, documents, databases, APIs and secure feeds
Extract
structured data pulled from messy inputs, with evidence
Enrich
identifiers resolved against reference data, attributes appended
Transform
mappings, calculations and normalisation
Validate
business rules, reconciliations, regulatory checks
Collaborate
exception queues, comments, review and approval
Deliver
XML, PDF, Excel, datasets and downstream systems
Worked example
A quarterly fund filing, end to end
A reporting team receives twelve files from three providers in four different layouts, plus two PDF statements. Previously this meant several days of copying, reconciling and checking. In Datox the same inputs run through a configured workflow.
What happens in the cycle
- Files land via SFTP and are matched to the reporting period automatically.
- Extraction produces structured holdings, exposures and investor records.
- Bare ISINs in the holdings file are resolved against reference data before provider classifications are mapped into the reporting taxonomy.
- Two exceptions and one low-confidence extracted value are routed to the reviewer.
- Once approved, the submission file and the internal evidence pack are generated together.
Reusability
Build once, run every reporting period
The first cycle is a configuration exercise: sources, mappings, rules, review steps and outputs. Every subsequent cycle reuses that configuration. New funds, entities or products are added onto the existing workflow rather than starting again.
- Period-over-period consistency, because the same logic is applied each time.
- Comparisons against prior periods can be used as a validation signal.
- Onboarding an additional entity or fund is configuration, not a new project.
- Changes are versioned, so you can see how the process evolved.
The problem
What Datox consolidates
Most reporting processes are assembled from parts that were never designed to work together: manual downloads, spreadsheet models, scripts maintained by one person, email approvals and last-minute checks. Datox replaces the gaps between those parts with a single controlled workflow.
- Manual data collection and re-keying across systems, files and documents.
- Spreadsheet models that hold critical reporting logic with no version history.
- One-off scripts and engineering tickets required for every reporting change.
- Checks performed by eye, late in the cycle, with no record of what was reviewed.
- Sign-off tracked in email threads instead of in the reporting workflow itself.
Core capabilities
What the platform does
Each capability is a part of the same platform, not a separate product. Configure once, then run the workflow every reporting period.
Data connectivity
Ingest Excel, CSV, PDF, Word, system exports, databases, APIs and SFTP feeds without forcing your sources into a template.
Learn moreData enrichment
Resolve identifiers against a Security Master and other governed reference data so sparse source records become reporting-ready before mapping and calculation.
Learn moreMapping & transformation
Map source fields to reporting fields, apply calculations and normalisation, and keep the logic visible and reusable.
Learn moreValidation & controls
Field-level rules, cross-field checks, reconciliations and thresholds that raise exceptions instead of silent errors.
Learn moreWorkflow & collaboration
Roles, ownership, comments, approvals and recurring cycles so reporting runs as a process, not a scramble.
Learn moreSource-to-output lineage
Trace any number in a finished report back through validation, transformation and extraction to the original source.
Learn moreSecurity & governance
Role-based access, audit trails, controlled AI usage and enterprise infrastructure practices.
Learn moreAI layer
AI agents do the repetitive work. Controls decide what is acceptable.
Datox agents read documents, propose mappings, apply transformations, run investigations and assemble outputs. They operate inside the workflow, against your rules, and they show their evidence. Deterministic logic, calculations, validations, thresholds, stays deterministic.
How the split works
- Agents handle extraction, mapping suggestions, anomaly investigation and output assembly.
- Rules engines handle calculations, validations and pass/fail decisions.
- Low-confidence or exception items are routed to a person, never quietly accepted.
- Every agent action is recorded with its inputs, evidence and outcome.
Ownership
Engineering integrates Datox. The business operates it.
Reporting logic changes constantly: a new fund, a revised mapping, an extra disclosure column, a different investor pack. Datox is configured so the reporting team can make those changes with the right permissions and an audit trail, instead of raising an engineering ticket for each one.
- Reporting owners configure mappings, rules and outputs within their permissions.
- Changes are versioned so you can see what changed, when and by whom.
- Engineering focuses on connectivity and infrastructure, not report maintenance.
- Workflows are reusable across periods, entities and funds.
Outputs
Any report. Any format.
The final step of the workflow produces the artefact your recipient actually needs, a regulator schema, a client-facing document, a working file or a dataset for downstream analytics.
Documents
PDF reports, factsheets, investor packs and board documents.
Working files
Excel and CSV outputs for teams that continue working with the data.
Datasets & feeds
Clean tabular outputs and API or file delivery into downstream systems and BI tools.