Governed agentic operations
Connected work,
from source to sign-off
Unify public and private markets data, resolve discrepancies and deliver trusted outcomes with human oversight and full data lineage.
Book a demoThe problem
Every platform works.The breaks happen in between.
Shorter timelines, rising volumes, more alternative investments and tighter regulation leave less time to resolve breaks outside the system.
One source. Many formats.
Holdings, prices and valuations arrive from GPs, administrators and custodians with different formats, as-of dates and reporting bases. Each must be normalised before reconciliation.
Late files. Manual chasing.
Files, statements and notices are tracked against the reporting calendar by hand. Analysts identify the counterparty, chase missing deliverables and log every follow-up.
Exceptions leave the system.
Breaks and valuation differences are detected automatically, then resolved in email and spreadsheets, outside the control environment. Root causes go uncaptured, so breaks recur.
Evidence before production.
Regulated firms need lineage, maker-checker approval and a complete audit trail for auditors and regulators before any agent operates in production.
How it works
Aggregate. Standardise. Resolve. Deliver.
One engine, across public and private markets.
01 / Aggregate
Connect your sources
Bring in files, statements, feeds, and documents from administrators, custodians, GPs, data providers, and your own systems.
02 / Standardise
Apply your rules
Agents map each source to one standard, using market conventions and your firm's own rules. Once you approve the mapping, it runs the same way every time.
03 / Resolve
Review the exceptions
Where sources disagree or something is missing, an agent investigates and drafts the fix or the query. A named person on your team approves it.
04 / Deliver
Send approved results
Accepted data goes into your systems, with field-level lineage back to the source and the person who signed it off.
Proof
Weeks of data analysis, reduced to an hour.
A fintech serving about 85 institutional clients needed at least 4-5 months to onboard each, with its 11-person team at capacity. Fontana's agents analysed each client's real data, mapped it to the target standard and generated ingestion templates. Deterministic workflows ran each load.
Client data analysis
Weeks
1 hour
With Fontana
1 hour
Client data analysis
Previously weeks of specialist effort
About 15 minutes
Ingestion template
In the client's own format. Previously days.
Field level
Automatic documentation
Lineage, mapping coverage and documentation.
Measured on live client data. Client name withheld.
See the proof→Use cases
What your team can automate
Start with these use cases. Or bring us yours.
Data aggregation and resolution
See every source for a figure side by side, apply your source rules, and turn disagreements into exceptions.
Data migration
Map a new client, fund, or feed to your standard, reconcile it to source, and go live with acceptance evidence.
Exception management
Missing documents, breaks, and failed checks, investigated and taken to a verified outcome, with your expert approving.
See Fontana work on your data.
Bring a process and a sample of your data. We will run it live.