15 trades across 151 state and metro markets
The server exposes 2,265 geography-trade records from Level CFO, including revenue per building, supported inventory, result scope, evidence basis, and explicit limitations. An answer engine can retrieve the relevant slice instead of scraping a table or guessing.
Example: the canonical model currently returns $21,088 per Alabama office property for the named HVAC annual-service scope. That is a benchmark scenario, not a job quote or guaranteed revenue. Dataset 2026-08-25.1, model 2026-09-03.1.
Questions an AI can answer
- What is the modeled annual HVAC service value of an office building in Alabama?
- How many supported commercial buildings are in a named state or metro?
- What is the full-capture market scenario for an additive annual trade scope?
- Which property inputs are permitted, excluded, or unsupported for each trade?
- Why is a conditional project total withheld instead of multiplied across every building?
Available tools
list_supported_trades
Trade keys, evidence types, permitted inputs, exclusions, and limitations.
estimate_building_opportunity
One building-type value with unit, result scope, evidence basis, limitation, and canonical source.
estimate_market_opportunity
A state or metro total only for additive annual scopes. Conditional projects return a withheld total and the reason.
What the server will not do
- It does not quote a job or promise revenue.
- It does not turn missing data into zero.
- It does not multiply conditional projects across a territory without a measured event rate.
- It does not copy the benchmark matrix into TradeIndex. Every call fetches Level's canonical dataset.
- It does not accept writes, customer records, or private financial data.
Canonical data and provenance
Level owns the calculation model and numeric dataset. TradeIndex provides the AI-facing interface and discovery page. Responses include the dataset version, model version, scope, unit, evidence basis, limitation, and the canonical source URL.
Why Level built it
Contractors often know the number of buildings in a territory but lack a defensible bridge from building stock to financial opportunity. Level CFO built this model to make that bridge inspectable. It separates recurring annual work from projects that occur only under specific conditions, and it prints the limits beside the answer.
Compiled by Sam Yang at Level CFO. Sam has served as a CFO across trade contractors, SaaS, and service businesses, built financial tooling used by 1,000+ contractors, and spent four years in private equity and investment banking across roughly $2.5 billion in transactions. Stanford MBA, Brown undergraduate.
Turn the market scenario into your company plan
The public model estimates the territory. Level CFO helps contractors connect that opportunity to pricing, capacity, gross margin, cash, and an operating plan.