A global asset manager running a multi-asset investment book, part-way through a strategic, multi-year programme to modernise its investment data platform. The data function spans portfolio management, operations, risk and regulatory reporting, and depends on a large and growing roster of market- and reference-data vendors.

Three years in, the platform programme had slowed to a crawl. New data sources arrived faster than the team could integrate them, and onboarding a new vendor was taking months. Governance and data lineage were maintained by hand and were perpetually incomplete — a serious problem given the firm’s regulatory reporting obligations. And when source schemas changed, integrations broke, so the operations team spent its days firefighting rather than advancing the platform. The programme was consuming effort without visibly moving forward, and stakeholder confidence was slipping.

The platform had been built, and was being run, largely through manual effort: large teams hand-documenting governance and lineage and maintaining a sprawl of integrations one at a time. That model does not scale with the number of data sources — every new vendor and every schema change generated another wave of manual work. Documentation could never keep pace with reality, so lineage was always slightly out of date, which is precisely the condition regulators are least forgiving of. The programme was not failing for lack of talent or investment; it was failing because the operating model was fundamentally manual.

DataSync introduced AI agents across the data function. The agents map incoming investment and vendor data to the platform’s model, generate governance and lineage documentation automatically, and maintain the business-rule and mapping knowledge in one place. When a vendor changes a schema, DataSync detects it and regenerates the affected mappings and documentation, so the operations team applies a reviewed update in hours instead of tracing the break for days. New-vendor onboarding became a largely automated analysis-and-mapping exercise rather than a bespoke project each time, and the accumulated knowledge made each subsequent source faster to add.

• New-vendor onboarding fell from months to weeks, clearing the backlog that had stalled the programme.
• Governance and lineage documentation became a by-product of the work rather than a separate, perpetually-behind effort — strengthening confidence in regulatory reporting.
• The manual maintenance burden dropped sharply, freeing the operations team to advance the platform instead of firefighting.
• Schema changes were caught and handled proactively, reducing the integration breakages that had dominated the team’s time.
• Overall programme delivery regained momentum, with a clear, compounding path as each new source was added.

DataSync is an AI platform for data migration and integration projects. Its agents analyse source systems, generate and maintain field-level mappings and transformation logic, surface data-quality issues, and produce the validated specification, governance and lineage documentation that engineering teams build from — turning work that once took months of manual analysis into weeks.

DataSync is a product of Blackstone DataSync Ltd, registered in England & Wales. To see how it could accelerate your next data project, visit datasync-ai.com or contact the team for a demonstration.