Learning path · Finance

Be the analyst who already knows AI.

You know the models, the ratios, the three-statement links. In 2026 that's the baseline — the hires who stand out are the ones who make AI do the grunt work and show up with the answer. This track turns you into that person, whether you're aiming for FP&A, a graduate analyst seat, or a finance internship.

Built for finance students & fresh grads. No fluff — concrete use-cases you can show an employer. Works on Claude, ChatGPT & Gemini.
Why this matters now

The 2026 finance desk expects it.

A finance grad who can drive AI does a day of modeling and analysis in an hour — and hiring managers know it. Two candidates with the same degree walk in; the one who can turn a messy actuals file into a clean forecast narrative in minutes is the one who gets the offer.

The shift

Grunt work is automated

Rekeying, formula-chasing, and cleaning exports are no longer where junior analysts prove themselves. The value moved up — to framing the question and explaining the answer.

What they screen for

"AI-capable" on the JD

FP&A and analyst postings increasingly list AI tooling alongside Excel. Showing you already work this way moves you past the pile that just lists coursework.

Your edge

Speed to insight wins

Nobody's impressed by a formula anymore. They're impressed by the analyst who ships the commentary the CFO can read — fast, correct, and in plain English.

The honest part: AI doesn't replace knowing finance — it rewards it. You still have to catch the wrong number. This track teaches you to be the human who does that at 10x speed.

The playbook

The AI skills that get you hired here.

Six things a finance hire can do with AI today. Each one is a concrete task you'd be handed in week one — and a concrete thing you can put on a portfolio. The prompt starters are real; steal them.

📐

AI-assisted financial modeling

Build the skeleton of a three-statement or unit-economics model in minutes, then audit it yourself. AI drafts the structure and flags where a link is likely broken; you own the logic.

Prompt → Draft a 3-statement model structure for a SaaS
business: revenue build from MRR, churn, and new logos.
List every driver, the formula for each, and 3 places
this kind of model usually breaks so I can check them.
🔮

Scenario & forecast prompts

Turn one base case into base / bull / bear in seconds. Ask for the assumptions behind each and the sensitivities that matter, so you walk into the review already knowing what breaks the plan.

Prompt → Here's my base forecast [paste]. Build a bull and
bear case. For each, state the 3 assumptions you changed,
by how much, and which single driver swings EBITDA most.
🖥️

Board-deck & commentary drafting

The part everyone dreads. Feed the variances, get a first draft of the CFO narrative — what moved, why, and the "so what." You edit for judgment instead of staring at a blank slide.

Prompt → Actuals vs budget attached. Write board commentary:
lead with the 3 biggest variances, explain the driver of
each in one sentence, and end with the action we're taking.
Tone: concise, exec-level, no hedging.
🔎

Large-dataset analysis

Point AI at a 50k-row GL export or transaction dump and find the story — outliers, duplicate vendors, the cost line that quietly doubled — without writing a pivot for every hunch.

Prompt → This is 12 months of expense lines. Find the top 5
categories by growth, flag any vendor paid more than twice
in a week, and list anything that looks like a miscoding.
💬

Explaining numbers to non-finance

Your real job is translation. Turn a variance bridge into a sentence a founder or department head actually understands — and get three versions for three audiences instantly.

Prompt → Explain why gross margin fell 4 points this quarter
to (a) the CEO in 2 lines, (b) the sales lead in terms of
what they control, (c) an investor. No jargon.
🧮

Excel + AI, together

Stop googling formula syntax. Describe the outcome and get the exact XLOOKUP / SUMIFS / dynamic-array formula, the Power Query step, or the reason your circular reference won't die.

Prompt → I need a formula that pulls the latest month's actual
for each account from a table where months are columns.
Give me the XLOOKUP, explain each argument, and a version
that returns "n/a" if the month is missing.
Your path

The route through the library.

Three volumes, in order. Start with the fundamentals everyone needs, add the productivity engine that makes finance work fast, then finish with the capstone that ties it to a CV line. Free to start.

  1. AI Foundations Free

    How AI actually works and how to talk to it so it gives finance-grade answers, not fluff. The base every other skill sits on.

  2. Productivity Core

    The engine for finance work — Excel formulas, large-dataset analysis, cleanup, and turning raw exports into decks and commentary in minutes. This is where the six skills above become muscle memory.

  3. JARVIS — the capstone Capstone

    Build an AI second-brain that holds your role's context — your metrics, your reporting calendar, your definitions — and drafts the analysis for you. The piece that makes "knows AI" real on your CV.

See the full catalogue and every volume on the courses page →

Prove it

Build one thing that gets you the interview.

Skills on a CV are claims. A project is proof. Here's one a finance student can build this week — it demonstrates every skill above and gives you something concrete to walk a hiring manager through.

Portfolio capstone

The Forecast Assistant

An AI workflow that takes a company's raw actuals and returns a board-ready narrative — the forecast, the variances, and the commentary a CFO could paste straight into a deck. Use any public company's financials or a mock dataset; the skill is the pipeline, not the numbers.

Step 1 · Ingest

Actuals in, cleaned

Feed messy actuals — a CSV or 10-K export. AI structures it, catches obvious miscodings, and builds the driver-based forecast off the trend.

Step 2 · Analyze

Variance + scenarios

Auto-generate actual-vs-plan variances and a base / bull / bear view, each with the assumptions that drive it spelled out.

Step 3 · Narrate

The board-ready story

Output the CFO commentary: top variances, the driver behind each, and the recommended action — in exec English, ready to present.

How to show it: record a 90-second screen walkthrough, or drop the before-file and after-narrative side by side. In an interview you say: "I built this — hand me your actuals and I'll do the same." That sentence lands harder than any bullet point.
Your move

Turn this into an offer.

Two steps that make you the AI-capable finance hire on paper and in practice: a CV tuned to your field, and the JARVIS capstone that proves you can actually do the work.