I spent a decade in legal tech, real estate, and enterprise sales before my MBA — watching smart people make critical decisions with inadequate tools. OPAL exists because institutional-quality analysis shouldn't require an institutional budget.
Before Kelley, I co-developed an AI SaaS product from concept to $45K MRR at a legal tech startup, rebuilt its revenue organization for 76% growth in 90 days, and led automation across sales, ops, and delivery.
At Zillow, I managed a $10M+ enterprise portfolio, navigated a company-wide platform pivot as an Accenture Root® certified facilitator, and coached a team overseeing $40M+ in transaction volume.
I also founded Modern Agent, a proptech consultancy that helped founders turn products into revenue — contributing to $4M+ in collective client funding and three NAR REACH accelerator placements.
The financial analysis industry has two tiers — and nothing in between.
Institutional terminals are comprehensive and reliable, and they are sold the way enterprise software is sold: by quote, to firms with a research budget and a procurement process. That shape does not fit a solo operator, an emerging fund manager, or a boutique consultancy.
General-purpose AI assistants are open to everyone and answer in conversational prose — which is a different job from producing a structured framework where every figure has to arrive attached to the evidence it came from.
OPAL fills the gap: ten MBA frameworks over licensed financial data and a fixed research allowlist, with claims citation-enforced, figures carrying a source you can open — the Defense sheet counts the ones that don’t — and the gaps named rather than filled.
Plans the run in a single fast call, then puts the researcher and the analyst to work in parallel and assembles what they return. It writes no analysis of its own.
Curates and cites the evidence harvest — a web search the Anthropic API restricts to a fixed allowlist of 28 domains before any query leaves: filings, the PR wires, analyst and research houses. It runs no search of its own and holds no calculator, and a citation that does not resolve to registered evidence never becomes a source link.
Reads licensed financial data that has already been registered as evidence. It has no calculator and no data feed: every figure it cites must already carry an id from that block. If the figure is missing, it says so rather than estimating.
Assembles the two narratives into one document — a single title, sections in a deliberate order, consistent formatting. It is a pure string transform with no model call, and it refuses to emit a report that has lost a citation tag.
Every company financial figure has to arrive already registered, carrying the id of the evidence it came from — the Analyst has no calculator and no data feed to invent one with. Links the model writes itself are stripped before anything renders, and the compile step refuses to emit a report that has dropped a citation tag. Where a figure is missing, the report says so.
SWOT, Porter's Five Forces, PESTEL, VRIO, and six more MBA frameworks. Claims are citation-enforced, and the Defense sheet counts the figures that carry no source.
Try It Free →A MEDDPICC deal scorecard, scored across all eight elements, with an AI coaching pass that reads your own notes and names the gaps your stage should have closed. It lives in the dashboard, one deal per account while it is in beta.
Open in the dashboard →Planned, not built: DCF modeling, WACC computation, Monte Carlo simulation, sensitivity analysis. The link joins the waitlist — that is all the page does today.
Join Waitlist →Disclaimer: The content provided by OPAL is for informational purposes only. Nothing on this site constitutes legal, tax, investment, financial, or other professional advice, nor does it constitute a solicitation, recommendation, or offer to buy or sell any securities. You assume sole responsibility for evaluating the merits and risks of any decisions made based on information provided through this platform.