AI training for finance teams, taught by practitioners who have done the work
A practitioner-led corporate program that takes your team from ad hoc prompting to reusable skills, working automations, and a governance standard, on the AI tools you have already deployed.
Built for finance teams
The core curriculum is the same. The exercises come from your function, on the deliverables your people already produce.
What your team gets
Every module ends with something built. The assets stay with your firm.
Prompting mastery
Consistently excellent output from the deployed tools, including from the messy inputs real finance work starts with: scanned PDFs, transcripts and multi-tab workbooks.
A skill library
Your firm's best work product, captured as reusable skills in your house formats. Written once by your best people, run by everyone.
Working automations
Recurring processes mapped and running end to end, no-code first, with human review where judgment belongs.
Finance use cases
Real finance tasks worked twice: once the manual way your team does it today, then AI-enabled, so the difference in time, quality and risk is visible side by side.
AI inside Excel and PowerPoint
Model auditing, formula generation, and commentary drafted against live numbers, in the tools where finance work actually happens.
An operating standard
A one-page policy on what goes into a model, what never does, and how output gets verified. Drafted with your team, ready for compliance.
The curriculum
Six modules, hands-on, taught on your team's real work. Open each one for detail.
1Foundations and tool fluencyHow the models work, at the depth needed to predict where they fail.
Deliberately short. Enough theory to choose the right model and anticipate its failures, then a working tour of your deployed stack. Organizations tend to overweight this part, so we do not.
- What a model is doing, in terms that predict its mistakes
- Hallucination, context limits and recency
- Choosing the right model and mode for the task
- Tour of the deployed stack, including finance-specific tools
2Prompting masteryStructure, decomposition, iteration, and the messy-input problem.
The difference between a mediocre answer and an excellent one is almost always the prompt. This module builds the habits on finance tasks, including the part most training skips: real work starts from ugly PDFs, scanned documents, transcripts and multi-tab Excel files, and handling those well is half of practical skill.
- Role, context, task, format: the anatomy of a working prompt
- Decomposition: breaking a memo into promptable parts
- The messy-input problem: scans, PDFs, transcripts, workbooks
- Exercise: rebuild a real deliverable from raw inputs
3Creating skillsTurning your best work into reusable assets the whole team can run.
A great prompt used once is a trick. Captured as a skill, it is an asset. This module codifies your firm's house formats, the IC memo, the variance narrative, the company profile, so output sounds like your firm regardless of who runs it.
- What makes a skill: instructions, examples, format, guardrails
- Golden examples: teaching the model from your best work
- Versioning, testing and improving skills over time
- Exercise: each team ships one working skill
4Finance-specific use casesExcel, documents, and the deliverables your function actually produces.
Finance does not live in a chat window. This module puts AI to work inside Excel and PowerPoint and across the documents your team handles every day, then runs specific use cases from your function twice: the manual way, then AI-enabled. The gap in time, quality and risk becomes the argument for adoption.
- AI-assisted model auditing, formula generation and explanation
- Commentary and narrative drafted against live numbers
- Document and deck workflows: drafting, redlining, versioning
- Use cases from your function, before and after: memos, screens, variance narratives, board materials
5Workflows and automationChaining skills into processes that run end to end.
Skills handle single deliverables. Workflows chain them: documents land, get classified and summarized, findings roll into the tracker, and a human reviews before anything moves. No-code first, with an optional technical extension for teams that have builders.
- Mapping a process into steps a model can run
- Where the human belongs: review gates and approvals
- No-code builds on your deployed platforms
- Exercise: one recurring process, automated and reviewed
6Verification and governanceYou own every number. AI drafts, you verify.
The module that makes the other five deployable. The standard is simple: you own every number, AI drafts and you verify, and juniors still learn to build from scratch before they automate. Then the rules of the road for finance: confidentiality, MNPI, model risk, and documentation that survives a compliance review. Taught by someone who has owned risk at the executive level.
- The verification standard in practice
- What never goes into a model, and how to work around it
- Confidentiality, MNPI and information barriers
- Deliverable: your team's one-page operating standard
Delivery formats
Scoped to your team, your tools and your calendar.
Executive briefing
For leadership. What the technology actually does, where the risk sits, and what a serious adoption plan looks like. Sets the mandate for everything below.
Intensive workshop
The six modules, hands-on, for a single team. First skills ship in the room and the team leaves with a working library.
Full program
The six modules across weeks, with build work between sessions so the skill library and automations mature against live work. Can be combined with Caserta's traditional finance training, including financial modeling, valuation and LBO modeling, for teams that want both built at once.
The platforms we train on
Hands-on across the four most finance organizations have deployed. Finance-specific and internal tools covered where your teams use them.
Enterprise deployments, custom GPTs, reasoning-heavy drafting and analysis.
Long-context document work, careful analysis, and skill-based workflows.
Deep research, Workspace integration, multi-modal analysis.
AI inside Excel, PowerPoint, Outlook and Teams.
Your instructors
Robert Monaco, MBA, CFA
Harvard MBA, CFA, CAIA, all four CPA sections passed. Former investment banker and private equity investor, founder of a venture-backed company that was raised and sold, and senior instructor to teams at companies and firms all over the world.
Scott Monaco, MBA
Wharton MBA. Former bulge bracket investment banker and venture capital investor. Senior analytics, product and risk leader at multiple venture-backed fintech companies through Series C and beyond.
Questions
Whichever your firm has deployed. We train hands-on across ChatGPT, Claude, Gemini and Copilot, plus finance-specific and internal platforms where your teams use them. We deliver on your approved stack rather than asking you to adopt something new.
It is built specifically not to. Every exercise starts with the team doing the task manually, then applying AI, then verifying the difference. Juniors learn to build from scratch before they automate, and the operating standard your team drafts makes that the house rule.
No. Exercises run on public filings and transcripts, or on sanitized versions of your own templates if you want output in your house style.
The people who produce the deliverables: analysts through directors in banking and investing seats, analysts through VPs in FP&A, corp dev and the CFO organization. The executive briefing is for the leadership group that sets policy.
By engagement, depending on format, team size and duration. Tell us what you are trying to build and we will come back with a scope and a quote.
Request a proposal
Tell us about the team, the tools you have deployed, and what you want your people able to do. We will come back with a recommended format and a scope.
Contact us about AI for Finance