Occam's Academy
A free, growing library of practitioner classes — grouped by level — on using the AI tools you already have to run your own forensic stock valuations. New courses roll out over the coming weeks, beginner first.
Practitioner-grade instruction on using Perplexity, Claude, and Gemini Flash to run real stock valuations on your own. Each course teaches one technique end to end — a research workflow, a valuation model, a way to resolve conflicting outputs.
A walkthrough of how Occam's Forensic Jury works internally. The methodology is proprietary; the Academy teaches the underlying skills so you can build your own analytical discipline.
Grouped by level. Start anywhere in Beginner — later levels assume the fundamentals from the earlier ones.
Teaching students to use Perplexity as a research co-pilot — without surrendering judgment to the machine. Covers the forensic mindset, the three lies of GAAP earnings, the Divergence Razor, the Cash-Based Rule of 40, the Zombie Death Clock, sector-adjusted valuation, and the Ticker-to-Thesis 60-minute workflow.
By the end, you can read a 10-K confidently, run a real ticker through the forensic discipline, and tell the difference between a company that earns its cash and a company that engineers it.
Tool: PerplexityA company is worth the cash it will produce. Build a discounted-cash-flow model from the ground up: the five building blocks — forecast free cash flow, discount rate, terminal value, present-value mechanics, and model structure — then a worked five-year DCF in a structured spreadsheet.
By the end, you can run a five-year DCF on a single company, source every input with Perplexity under cite-or-verify discipline, and defend each assumption you used to get there.
Tool: PerplexityStep beyond a stated discount rate. Use Claude to derive the weighted average cost of capital, then run a reverse DCF to see the growth the market is already pricing in. Revenue growth assumptions, terminal value, sensitivity tables — and where each one hides its weak points.
By the end, you can defend your discount rate from first principles and read what a stock price implies about the future.
Tool: ClaudeWhen your DCF says one thing, peer comparison says another, and the qualitative read says a third — you have a verdict problem, not a math problem. Learn to use Gemini Flash as an independent arbiter that can stress-test conflicting outputs, surface which assumption is doing the load-bearing work, and write a tight summary you would be willing to defend.
By the end, you can resolve cross-model disagreement instead of papering over it.
Tool: Gemini FlashFuture levels expand into 401-series workflows: sector-specific routing, earnings-call hedging detection, kill-criteria automation, and applying the discipline at scale. Subscribe to the newsletter on the homepage to be notified as each course launches.