Invitation-only research workbench

FinBro, agentic AI value research with institutional grade data.

Company valuations, stock screens, and commodity and Bitcoin analysis built on fundamentals. Add an optional market thesis to a new valuation or commodity run without steering its conclusion, or revisit an eligible saved report after a complete independent current analysis. FinBro is a directable research workspace, not a one-shot prompt box: add notes, ask follow-ups, and inspect the evidence categories and available source and date context, which vary by report format, workflow, and run.

By the numbers

What a typical FinBro run looks like.

Typical runtime and usage for core workflows, followed by the coverage built into the research workspace.

Company valuation
25 minTypical end-to-end runtime
Stock screen
40 minTypical multi-name runtime
Lite run
~$30Typical usage charge · billed per token

Typical figures, not service guarantees. Runtime varies by request; usage varies with input and output tokens. The platform subscription is separate.

Sector lenses
13
Grounded evidence categories
8
Projection horizons
2Forward Projected Value · 3yr · 5yr
History
Annual + quarterlyAvailability varies by listing, period, and field
Example output from the workbench

Examples of FinBro's analysis output.

Illustrative examples of what a FinBro analysis contains — each with its model estimate and the evidence basis behind it.

These are illustrative examples: the companies are fictional and no real instrument is valued. Figures are model estimates — not price forecasts, guarantees, or a record of investment returns. Capital at risk.

Company analysis

Example Co. A - EXA

Illustrative, fictional example

Modeled base intrinsic value
$360
Margin of safety (model)
30%

For this fictional company, the model estimated a base intrinsic value of $360 per share and a 30% margin of safety (model) against the observed market price, based on company fundamentals.

Evidence basisCompany fundamentals

Commodity analysis

Bitcoin - BTC

Illustrative example

Valuation state
No intrinsic-value model for this asset class

A Bitcoin analysis examines the macro drivers and their scenario implications. No intrinsic-value model exists for this asset class, so no valuation state or margin of safety is computed.

Evidence basisMacro data, US rates, inflation and adoption

Company analysis

Example Co. B - EXB

Illustrative, fictional example

Modeled base intrinsic value
$150
Margin of safety (model)
15%

For this fictional company, the model estimated a base intrinsic value of $150 per share and a 15% margin of safety (model) against the observed market price, based on company fundamentals.

Evidence basisCompany fundamentals

These examples illustrate what an analysis contains. They are not a record of FinBro output and not realized investment performance; no return, hit rate, benchmark result or backtest is shown.

This describes the relationship between the observed market price and FinBro's modeled intrinsic-value range. It is not a recommendation to buy, sell or hold, is not based on individual circumstances, and does not assess suitability. Model outputs are estimates, not forecasts or guarantees.

The process

One question, many minds, one answer.

Specialists work in parallel on independent dimensions of the thesis, then their findings are challenged and reconciled into a single view.

Relevant specialist lenses
MacroCompanySectorValue

Product overview: Request → Analysis → Challenge → Report → Follow-up

The route depends on the task. Company Valuation and Commodity & Bitcoin Analysis each use their own complete current-analysis path. Revisit thesis runs the complete current-analysis path appropriate to the selected eligible report before comparing it with that report. Stock Screening uses its own screening path; after reviewing the screen results, you can open any listed instrument in a separate Company Valuation.

  1. 01

    Ask

    Start with a plain-language question or workflow input, or select an eligible saved report to revisit. No query syntax or prompt is required for a saved-report revisit.

  2. 02

    Parallel specialist analysis

    The relevant specialist lenses build the current analysis independently from the available evidence. A selected prior report or optional market premise does not set the conclusion.

  3. 03

    Adversarial challenge

    A dedicated stage argues the other side — disconfirming evidence, historical analogues, hidden assumptions and second-order effects — and flags gaps by severity. A critical gap pauses the run for your decision.

  4. 04

    Board synthesis

    Independent frontier models review the completed analysis and reconcile it into a single second-opinion view — depth without a single black box.

  5. 05

    Sourced report + PDF

    A long-form, cited write-up is streamed live and exportable as a PDF. An eligible new valuation or commodity run may append a labeled Scenario Assessment after its independent conclusion; a saved-report revisit instead leads with an after-the-fact comparison and retains the complete current analysis.

Learn how it works
Request
What you can run

Four ways to put the workbench to work.

Single-name analysis

Company Valuation

When you already have a company in mind and want a full, sourced view of its value.

Input
Ticker, company name, or a question about one business — plus, optionally, a market thesis to test against the analysis.
Process
Parallel macro, company, sector and valuation lenses, followed by an adversarial challenge and synthesis.
Output
A sourced report with the valuation state, modeled intrinsic-value range, margin of safety (model), scenarios, peer context, risks, 3-year/5-year Forward Projected Values and PDF export.
Find candidates first

Stock Screening

When you want to see which instruments meet documented screening criteria before you choose what to study.

Input
Universe (US, EU or both), market cap, sector, country and candidates analysed, plus optional free text read by a fixed keyword table.
Process
Instruments that meet the documented listing and data-availability eligibility checks are listed in market-cap order; each listed instrument is then analysed for valuation state, margin of safety (model), quality and growth.
Output
Screen results ordered by market cap; run a separate Company Valuation on any listed instrument.
Macro asset view

Commodity & Bitcoin Analysis

When you want a fundamentals-first read on a commodity or Bitcoin — metals, energy, agriculture or the digital asset — rather than a single equity.

Input
A commodity or Bitcoin — a metal, an energy or agricultural commodity, or the digital asset — plus, optionally, a market thesis to test against the analysis.
Process
Macro-regime, supply/demand and cost-curve lenses, followed by an adversarial challenge and synthesis.
Output
A sourced analysis with illustrative 3-year model scenario bands, the macro drivers behind them, and PDF export.
Prior thesis, current evidence

Revisit thesis

Revisiting an eligible saved Company Valuation, Commodity & Bitcoin Analysis, or prior Revisit thesis report against a current full analysis.

Input
One eligible saved report owned by the signed-in user. Stock-screen reports are ineligible, and no prompt is required.
Process
The complete current analysis runs independently of the selected report; comparison happens only afterward.
Output
One sourced report leading with the prior-versus-current comparison and including the complete current analysis. Individual comparison dimensions may be unavailable.

Test your thesis

Optionally add a market thesis to a new Company Valuation or Commodity & Bitcoin Analysis. FinBro reaches its conclusion independently, then may append a labeled Scenario Assessment showing where available evidence supports or challenges the premise and how the picture could shift if it held; the assessment is unavailable when the evidence cannot assess it.

Read the docs
Analysis
Why it is different

A rules-driven pipeline around deep AI analysis.

FinBro is not an open-ended prompt that improvises its own process. The system applies repeatable evidence checks across data, modelling, business inspection and value-investing synthesis before it produces a final view.

See the operating process
01

Rules-driven pipeline

Every run follows the same evidence checklist before synthesis, so the process is orchestrated rather than improvised.

02

Thousands of data points

Financial statements, filings, macro series, ownership signals, market inputs, valuation context and contract-award records are reconciled into one evidence base.

03

Financial modelling layer

Valuation methods, scenario models, peer context and margin-of-safety checks turn raw inputs into a valuation you can inspect.

04

Business and leadership inspection

Company quality, sector dynamics, business model, competitive position and leadership signals are reviewed where public evidence supports them.

05

Value-investing synthesis

The final view weighs quality, growth potential, intrinsic value, margin of safety (model), thesis risk and Forward Projected Values instead of stopping at a surface summary.

Analysis
Grounding

Inspect how analysis evidence is classified.

See how evidence categories, date bases, attribution practices, and limitations vary across workflows and runs.

  • Public record
  • Reported/aggregated
  • Market
  • Estimate
  • Derived
Inspect coverage, freshness, and limitations
US + EUequities
Commoditiesmetals · energy
Bitcoindigital asset
Representative categories
Financial statements & ratios - Reported/aggregated + derived
Freshness basisFiling/reporting period
Market & price data - Market
Freshness basisMarket as-of date
Valuation inputs & outputs - Derived
Freshness basisCalculated from dated inputs
AnalysisStock Screening branch
Screen results

A neutral screen, in a documented order.

The stock screener lists the instruments that meet documented screening criteria — your universe, market cap, sector and country, plus listing and data-availability eligibility checks — ordered by market cap. Once listed, each analysed instrument shows its valuation state, margin of safety (model), quality and growth; run a separate Company Valuation on any one for the full write-up.

Universe · market cap
US · EU · both · any size to mega
Sector · country
Fixed lists, optional
Candidates analysed
Coverage cap, 1–20
Open the screener guide
Illustrative workflow
No live data; fictional instruments.
ordered by market cap
  • Instrument A
    market cap $48B
    Industrials
    Market price below the modeled intrinsic-value range

    Meets the listing and data-availability eligibility checks; analysed within the budget.

  • Instrument B
    market cap $21B
    Financials
    Market price within the modeled intrinsic-value range

    Meets the listing and data-availability eligibility checks; analysed within the budget.

  • Instrument C
    market cap $9B
    Retail
    Insufficient evidence for a modeled range

    Analysed, but the available inputs do not support a modeled range.

Instruments are listed because they meet the documented screening criteria, ordered by market capitalisation. Listing is not a selection or recommendation; instruments outside the analysis budget were not evaluated.

This describes the relationship between the observed market price and FinBro's modeled intrinsic-value range. It is not a recommendation to buy, sell or hold, is not based on individual circumstances, and does not assess suitability. Model outputs are estimates, not forecasts or guarantees.

Challenge
Challenge gate

The thesis has to survive the other side.

FinBro does not only assemble the positive case. Before synthesis, it looks for what would weaken, delay, reverse or reprice the thesis, then keeps those findings visible.

Critical gap decision
Refocus
Ask for a tighter run that closes the gap.
Proceed
Keep the caveat visible in the final report.
Stop
Do not turn a broken thesis into a confident answer.
Read the challenge flow
01

Disconfirming evidence

Looks for facts that weaken the thesis instead of only supporting it.

02

Historical analogues

Checks whether similar setups ended differently than the headline story suggests.

03

Hidden assumptions

Calls out the growth, margin, rate, cycle or execution assumptions the case depends on.

04

Second-order effects

Tracks what can happen after the obvious first move — competitors, regulation, capital costs and customer behavior.

Report
What you get back

A report that reads like a memo, not a chat transcript.

A long-form, sectioned report — valuation state, scenarios, peers, risks, challenge findings, Forward Projected Values and sources. It streams live as the analysis runs and exports as a PDF. Report format, source and date context, and citation coverage vary by workflow and run.

See the valuation guide
report anatomy
  • 01Thesis & summaryThe valuation state up front — market price below, within or above the modeled intrinsic-value range, or insufficient evidence — with the core argument in a few lines.
  • 02Valuation scenariosBear, base and bull cases built with the valuation method that fits the business, assumptions made explicit, giving the modeled intrinsic-value range and margin of safety (model).
  • 03Macro & sector contextWhere the company sits in the current regime — rates, cycle, and the sector dynamics that move it.
  • 04Peer comparisonHow the name stacks up against comparable companies on the metrics that matter for its sector.
  • 05Thesis challengesThe adversarial stage's findings — what could break the case, ranked by severity — addressed rather than buried.
  • 06Forward Projected Value3-year and 5-year model scenario values per share in USD (with EUR conversion where relevant) — modeled values, not forecasts.
  • 07SourcesSource categories and, when available, filing, reporting-period, series, estimate-date, or market as-of context; citation coverage may vary by run.
Follow-up
The workspace

A research workspace, not a one-shot prompt box.

Continue in the workspace: add notes, ask follow-up questions, inspect the report beside the conversation, or separately start Revisit thesis from an eligible saved report.

  • Plain-language requests
  • Live progress labels
  • Text-file context
  • Report drawer + PDF
  • Saved screener defaults
  • Fresh run each time
  • Test your own thesis
  • Revisit saved reports
Who it's for

Built for the self-directed value investor.

Deep fundamental analysis used to require an analyst team and expensive data terminals. FinBro puts multi-method valuations, scenario and margin-of-safety work, peer and sector context, an adversarial challenge and a multi-model second opinion within reach of individual investors, small and mid-size firms, and family offices — the same discipline, a fraction of the manual work.

Individual investorsSmall & mid-size firmsFamily offices
The investor

FinBro is for the investor who wants to look past the index and understand individual businesses — across equities and commodities. When the market stress-tests your thesis, you want to have already stress-tested it yourself.

Your desk

Your desk, not a black box. No analyst team to hire, no private banker reading someone else's research to you, no five-figure data terminal. FinBro is the technical workbench that does the fundamental heavy lifting — you bring the thesis and the sectors you know, and every finding comes back sourced and open to inspection, never 'just trust us.'

General research information addressed to the public: model estimates and opinions about value. It is not a recommendation to buy, sell or hold, is not based on individual circumstances, and does not assess suitability. Capital at risk.

Invitation only

Built for deep fundamental analysis.