What the project actually gives you

The project packages reference agents, vertical plugins, skills, slash commands, MCP connector settings, and managed-agent cookbooks for common financial-services workflows.

Its value is the structure: named agents for work like pitch creation, market research, model building, valuation review, GL reconciliation, month-end close, statement audit, and KYC screening.

• Agent templates for end-to-end workflows such as pitch, meeting prep, market research, earnings review, and KYC screening.
• Vertical bundles for investment banking, equity research, private equity, wealth management, fund administration, and operations.
• Managed Agent cookbooks for teams that need to embed the same prompts and skills behind their own workflow engine.

Where the commercial value appears

The strongest pilots usually reduce the time analysts spend turning source material into a reviewable first draft. That includes comps, DCF assumptions, earnings notes, IC memo sections, reconciliation narratives, and deck quality checks.

The project is deliberately conservative. It stages output for human review instead of making investment recommendations, posting to ledgers, or approving onboarding decisions.

How to scope the first rollout

Start with one high-frequency workflow and one operations workflow. A common pairing is earnings review plus GL reconciliation, because the team can measure both research-cycle speed and exception-handling quality.

Before launch, define allowed data sources, reviewer ownership, output templates, retention rules, and the point where the agent must stop and ask for approval.

Common questions