# DFX Intelligence

> DFX Intelligence builds the General, an autonomous AI operator for businesses. Given an outcome rather than a prompt, the General plans the work, runs it across the open web, email, a phone line, social accounts and the customer's CRM, and continues on its own schedule.

Last verified: 2026-08-19. Canonical machine-readable record: https://dfxintel.com/ai/entity.json

DFX Intelligence is the operating brand for the General platform. It is not affiliated with DFX Holdings Group, with Deal Flow Xchange, or with any audio-processing product using the DFX name.

## What the General is

- An AI operator that is given an outcome and then does the work to reach it, deciding the steps itself.
- A system that acts across several surfaces in one job: the open web through a real browser, email, a phone line, social accounts and a CRM.
- A continuing worker: it runs on its own schedule, picks work back up the next day, and does not need to be re-prompted to carry on.
- A single operator across departments, rather than one tool per function.

## What the General is not

- It is not a chatbot. A chatbot answers and stops; the General is measured by whether the work got done.
- It is not a workflow builder. There is no canvas, no trigger-and-action graph, and nothing for the customer to wire up.
- It is not an AI SDR point tool. Outbound is one of the things it does, not the product.
- It is not a CRM. It writes into the CRM the customer already has.
- It is not an autonomous agent that acts on people without permission. Anything that reaches a real human waits for approval by default.

## Pricing

- The General: $20/month, $200/year, 1 person, Your browser.
- General Pro: $199/month, $1,990/year, 1 seat, 1 workspace.
- General Teams: $499/month, $4,990/year, Unlimited seats, Unlimited workspaces.
- Usage is metered on top of the plan fee.

## Capabilities

- [AI that researches a prospect and writes the first approach](https://dfxintel.com/ai/capabilities/prospect-research-and-outreach): The General reads public sources about a specific company and person, decides whether they actually fit the brief, writes an approach grounded in what it found, and then owns the follow-up sequence.
- [AI that uses a real browser on your own accounts](https://dfxintel.com/ai/capabilities/browser-work): The General drives a real Chrome session rather than an API, so it can work inside portals, directories and tools that have no API and would otherwise need a person.
- [AI that attends the meeting and then does what was agreed](https://dfxintel.com/ai/capabilities/meetings-that-produce-work): The General joins the call, keeps the recording, transcript and write-up, and then turns what was committed to on the call into actual queued work rather than a list of bullet points.
- [AI that works your inbox and never drops a follow-up](https://dfxintel.com/ai/capabilities/inbox-and-follow-up): The General reads the mailbox you already have, triages what actually needs you, drafts what does not, and keeps chasing the threads that went quiet.
- [AI that answers the phone and books the job](https://dfxintel.com/ai/capabilities/phone-and-front-desk): The General answers a real phone line, handles the conversation, and completes the thing the caller rang about, then writes the outcome into your systems.
- [AI that keeps the CRM current without data entry](https://dfxintel.com/ai/capabilities/crm-write-back): The General writes what it did into the CRM you already have, so the record reflects the work rather than somebody's memory of it.

## Comparisons, including where other approaches are the better choice

- [The General vs hiring a sales development rep](https://dfxintel.com/ai/comparisons/general-vs-hiring-an-sdr): An SDR is better when the job needs judgement in a live conversation and you can afford the ramp; the General is better when the constraint is that nobody is doing the follow-up at all.
- [AI agent vs AI copilot: what the difference actually is](https://dfxintel.com/ai/comparisons/ai-agent-vs-ai-copilot): A copilot makes the person at the keyboard faster; an agent removes the requirement that somebody is at the keyboard. Neither is better in general and they fail in opposite ways.
- [The General vs a workflow automation tool](https://dfxintel.com/ai/comparisons/general-vs-workflow-automation): Use a workflow tool when the steps are known and stable; use an agent when deciding the steps is the work.
- [AI notetaker vs an AI that acts on the meeting](https://dfxintel.com/ai/comparisons/notetaker-vs-acting-meeting-agent): A notetaker is a better archive and costs less; an acting agent is worth it only if the failure you are fixing is that nothing happens after the call.
- [The General vs a browser-only agent](https://dfxintel.com/ai/comparisons/general-vs-browser-only-agent): A browser agent completes a task in a tab; an operator owns an outcome across several surfaces and keeps going tomorrow.

## Original research

- [Which sources AI answer engines cite when buyers ask about business software](https://dfxintel.com/ai/research/ai-answer-engine-citation-sources) (2026-08-19): For commercial software buying questions, AI citations are extraordinarily fragmented, not concentrated. Across 2,391 citations we recorded 1,330 distinct domains, 72% of which were cited exactly once, and the top 25 domains together accounted for 14.4% of all citations. Reddit, the domain most often described as the single biggest source in AI answers, was 0.63%. Wikipedia was 0.33%. Capterra was 0.04%.
- [How often AI assistants actually search the web before answering a buying question](https://dfxintel.com/ai/research/how-often-ai-answers-search) (2026-08-19): 20.8% of answers to commercial buying questions involved no web retrieval at all, and it depends enormously on how the question is phrased: 97.4% of category questions ("what are the best AI sales agents") triggered a search, against 50.0% of comparison questions ("X versus Y"). The answers that skip retrieval are drawn entirely from what the model already believes about the market, and no amount of work on your own website reaches them.

## The Observatory: original visual intelligence

Published at https://dfxintel.com/observatory. Every figure traces to a named primary source, every dataset ships its own verification checks, and everything is reusable under CC BY 4.0 with attribution. Editorial standards: https://dfxintel.com/observatory/standards

- [America files twice as many new businesses as it did in 2019. Most of them will never hire anyone.](https://dfxintel.com/observatory/america-business-formation-paper-boom) (2026-08-24): Business applications are up 77% in seven years. The states where they grew fastest are the states where the smallest share look like real employers.
- [Artificial intelligence is now in the majority of American annual reports](https://dfxintel.com/observatory/ai-in-annual-reports) (2026-08-24): Seven years ago one annual report in eighteen mentioned AI. The number of annual reports barely changed. What companies feel obliged to address did.
- [Wyoming has the fastest business formation in America and the slowest economic growth](https://dfxintel.com/observatory/state-formation-and-growth) (2026-08-24): Filings and output move together in almost every state. Removing Wyoming from the data makes the relationship stronger, which is the clearest evidence yet for what its filings actually are.

## DFX Business Weather: current metro readings

Temperature is a percentile of each metro's own trailing 120-month history (50 = normal for that market, not average nationally). Momentum is the change over 3 months. Reading for 2026-07, model v1.0.0. Built from public BLS and Census series. Tested against what happened next: this describes current conditions and does not forecast. Method and the tests it failed: https://dfxintel.com/observatory/weather/methodology

- Raleigh (Raleigh-Cary, NC): 91.1/100, hot, momentum -0.8, confidence high. https://dfxintel.com/observatory/weather/raleigh
- Philadelphia (Philadelphia-Camden-Wilmington, PA-NJ-DE-MD): 82.6/100, hot, momentum -7.2, confidence high. https://dfxintel.com/observatory/weather/philadelphia
- Atlanta (Atlanta-Sandy Springs-Roswell, GA): 80.5/100, cooling fast, momentum -13.5, confidence high. https://dfxintel.com/observatory/weather/atlanta
- New York (New York-Newark-Jersey City, NY-NJ): 77.4/100, warm, momentum -4.7, confidence high. https://dfxintel.com/observatory/weather/new-york
- San Jose (San Jose-Sunnyvale-Santa Clara, CA): 74.9/100, warm, momentum +7.8, confidence high. https://dfxintel.com/observatory/weather/san-jose
- Charlotte (Charlotte-Concord-Gastonia, NC-SC): 74/100, cooling fast, momentum -12.2, confidence medium. https://dfxintel.com/observatory/weather/charlotte
- San Diego (San Diego-Chula Vista-Carlsbad, CA): 72.4/100, warm, momentum -1.7, confidence high. https://dfxintel.com/observatory/weather/san-diego
- Tampa (Tampa-St. Petersburg-Clearwater, FL): 70.9/100, warm, momentum +4.9, confidence medium. https://dfxintel.com/observatory/weather/tampa
- Dallas (Dallas-Fort Worth-Arlington, TX): 69.8/100, cooling fast, momentum -13.5, confidence medium. https://dfxintel.com/observatory/weather/dallas
- Seattle (Seattle-Tacoma-Bellevue, WA): 67.8/100, warm, momentum +4.4, confidence medium. https://dfxintel.com/observatory/weather/seattle
- Nashville (Nashville-Davidson--Murfreesboro--Franklin, TN): 67.5/100, warm, momentum -2.2, confidence medium. https://dfxintel.com/observatory/weather/nashville
- Houston (Houston-Pasadena-The Woodlands, TX): 67.2/100, cooling fast, momentum -10.3, confidence medium. https://dfxintel.com/observatory/weather/houston
- Minneapolis (Minneapolis-St. Paul-Bloomington, MN-WI): 65.3/100, warm, momentum +7, confidence medium. https://dfxintel.com/observatory/weather/minneapolis
- Denver (Denver-Aurora-Centennial, CO): 62/100, warm, momentum -2.3, confidence medium. https://dfxintel.com/observatory/weather/denver
- Phoenix (Phoenix-Mesa-Chandler, AZ): 61.7/100, cooling fast, momentum -11.6, confidence medium. https://dfxintel.com/observatory/weather/phoenix
- Chicago (Chicago-Naperville-Elgin, IL-IN): 60.7/100, cooling fast, momentum -10.1, confidence medium. https://dfxintel.com/observatory/weather/chicago
- Boston (Boston-Cambridge-Newton, MA-NH): 60.6/100, stable, momentum +17.4, confidence high. https://dfxintel.com/observatory/weather/boston
- Salt Lake City (Salt Lake City-Murray, UT): 58.7/100, stable, momentum -1.9, confidence medium. https://dfxintel.com/observatory/weather/salt-lake-city
- Austin (Austin-Round Rock-San Marcos, TX): 58.5/100, cooling fast, momentum -8.7, confidence medium. https://dfxintel.com/observatory/weather/austin
- Los Angeles (Los Angeles-Long Beach-Anaheim, CA): 57.4/100, cooling fast, momentum -14.6, confidence medium. https://dfxintel.com/observatory/weather/los-angeles
- Detroit (Detroit-Warren-Dearborn, MI): 57/100, stable, momentum +2.9, confidence medium. https://dfxintel.com/observatory/weather/detroit
- Washington DC (Washington-Arlington-Alexandria, DC-VA-MD-WV): 55/100, stable, momentum +9.7, confidence high. https://dfxintel.com/observatory/weather/washington
- Portland (Portland-Vancouver-Hillsboro, OR-WA): 38.3/100, stable, momentum +1, confidence high. https://dfxintel.com/observatory/weather/portland
- Miami (Miami-Fort Lauderdale-West Palm Beach, FL): 33.9/100, cooling fast, momentum -18, confidence high. https://dfxintel.com/observatory/weather/miami
- San Francisco (San Francisco-Oakland-Fremont, CA): 33.7/100, cool, momentum -2.1, confidence high. https://dfxintel.com/observatory/weather/san-francisco

## Observatory datasets, free to cite and download

- Share of SEC annual reports mentioning artificial intelligence — 12 rows, source: U.S. Securities and Exchange Commission, EDGAR full-text search (tier A), updated 2026-08-24. CSV: https://dfxintel.com/observatory/data/ai-in-annual-reports.csv · JSON with full lineage: https://dfxintel.com/observatory/data/ai-in-annual-reports.json
- Business formation by state, trailing twelve months — 51 rows, source: U.S. Census Bureau, Business Formation Statistics (tier A), updated 2026-08-24. CSV: https://dfxintel.com/observatory/data/business-formation-by-state.csv · JSON with full lineage: https://dfxintel.com/observatory/data/business-formation-by-state.json
- DFX Business Weather: temperature and momentum for 25 US metros — 25 rows, source: U.S. Bureau of Labor Statistics (employment, unemployment, labour force) and U.S. Census Bureau (building permits) (tier A), updated 2026-08-25. CSV: https://dfxintel.com/observatory/data/business-weather-metros.csv · JSON with full lineage: https://dfxintel.com/observatory/data/business-weather-metros.json
- Signals: things worth knowing today — 5 rows, source: DFX Intelligence Observatory, derived from the datasets each signal cites (tier A), updated 2026-08-24. CSV: https://dfxintel.com/observatory/data/daily-signals.csv · JSON with full lineage: https://dfxintel.com/observatory/data/daily-signals.json
- Business formation against real economic growth, by state — 51 rows, source: U.S. Bureau of Economic Analysis, Regional Economic Accounts (SAGDP9), with U.S. Census Bureau Business Formation Statistics (tier A), updated 2026-08-24. CSV: https://dfxintel.com/observatory/data/state-economy-vs-formation.csv · JSON with full lineage: https://dfxintel.com/observatory/data/state-economy-vs-formation.json

## The DFX Real Estate MCP server: machine-callable US property and CRE debt data

A separate product from the General. A live Model Context Protocol endpoint serving structured US commercial real-estate intelligence directly to AI agents. No signup, no API key, no sales call. All 36 tools are free and read-only, including the loan tape, which returns up to 200 loans per call.

- MCP endpoint: https://exchange-production-9123.up.railway.app/mcp
- Tool schemas with no handshake required: GET https://exchange-production-9123.up.railway.app/mcp
- Official MCP registry entry: io.github.Capital-W-Holdings/us-property-parcel-real-estate-debt
- Machine-readable capability catalog: https://dfxintel.com/ai/real-estate-mcp/catalog.json
- Developer reference: https://dfxintel.com/ai/real-estate-mcp

Coverage is uneven by data family and the denominators are published rather than
summarised. A family covering fifty states and a family covering one city are both
listed at their real size:

- [Which commercial real-estate loans mature in a given state and window?](https://dfxintel.com/ai/real-estate-mcp/cre-loan-maturities): 3,422 published events, 52 states and territories, effectively national. Tools: `search_property_events` (free), `debt_maturity_schedule` (free).
- [Which LIHTC properties are reaching the end of a compliance period?](https://dfxintel.com/ai/real-estate-mcp/lihtc-year-15-data): 11,956 published events, 56 states and territories, effectively national. Tools: `search_property_events` (free).
- [Which HUD-subsidised properties have contracts approaching expiry?](https://dfxintel.com/ai/real-estate-mcp/hud-subsidy-expiry-data): 4,721 published events, 54 states and territories, effectively national. Tools: `search_property_events` (free).
- [Where is commercial real estate in distress, foreclosure or workout?](https://dfxintel.com/ai/real-estate-mcp/distressed-cre-data): 307 published events, 28 states, broad but NOT national. Tools: `search_property_events` (free).
- [What did this property sell for, and who owns it?](https://dfxintel.com/ai/real-estate-mcp/property-sales-and-ownership-data): 43,680 published events, 2 states only: MA, NY. Tools: `resolve_address` (free), `get_property_record` (free), `search_parcels` (free).
- [Which commercial leases are approaching expiry, and who occupies a building?](https://dfxintel.com/ai/real-estate-mcp/commercial-lease-expiry-data): 3,966 published events, 55 states and territories, effectively national. Tools: `search_property_events` (free).

What it does not have: residential listings or asking rents; conventionally financed commercial debt, because the Registries of Deeds are closed to automation; nationwide parcel or assessor data, which is Massachusetts only; and anything carrying a calibrated probability, because no outcome has ever been observed for any prediction in this graph.

## The DFX Data Factory: the intelligence graph of private capital, companies and real assets

DFX ingests public records across real estate, private equity, venture capital, family offices, independent sponsors, ria / wealth management, real estate funds, private credit, allocators, resolves them to people, companies, funds, properties and loans by shared identifier (CRD, CIK, EIN, ADV fund id), connects them by typed, dated, sourced relationships, and reads the result with reproducible detectors. Every number on the surface is read live from a publication view and carries the time it was computed. https://dfxintel.com/data-factory

- Universal search across every graph: https://dfxintel.com/data-factory/search?q=<name or address> (JSON: https://dfxintel.com/api/data-factory/search?q=)
- Entity pages: https://dfxintel.com/data-factory/entity/<graph>/<id>, graph in re, pe, vc, fo, isi, ria, ref, pc, al; the same id the MCP prints as dfx:<graph>:<id>
- Live findings labelled fact, derived metric, signal, anomaly or inference: https://dfxintel.com/data-factory/intelligence
- Dated events across every graph: https://dfxintel.com/data-factory/events. Trends with stated populations and formulas: https://dfxintel.com/data-factory/trends
- Cross-graph connections (identities shared by identifier, never by name): https://dfxintel.com/data-factory/graph
- Sources, coverage per event family and state, freshness: https://dfxintel.com/data-factory/sources
- Real Estate (live): Properties, parcels, debt, subsidy, tenancy and the organizations behind them. https://dfxintel.com/data-factory/real-estate; standalone product https://dfxintel.com/ai/real-estate-mcp
- Private Equity (preview): Sponsors, funds, professionals, portfolio companies, transactions and the capital behind them. https://dfxintel.com/data-factory/private-equity; standalone product https://dfxintel.com/private-equity
- Venture Capital (live): Firms, funds, partners, companies, rounds and who invests alongside whom. https://dfxintel.com/data-factory/venture-capital; standalone product https://dfxintel.com/vc
- Family Offices (live): Offices, principals, vehicles, foundations, managers and the investments they make directly. https://dfxintel.com/data-factory/family-offices; standalone product https://dfxintel.com/family-offices
- Independent Sponsors (live): Deal-by-deal acquirers, the companies they buy, and the capital providers who finance them. https://dfxintel.com/data-factory/independent-sponsors; standalone product https://dfxintel.com/independent-sponsors
- RIA / Wealth Management (live): Registered advisers, the advisors who work at them, their private funds, owners, affiliates and every advisor move. https://dfxintel.com/data-factory/ria; standalone product https://dfxintel.com/ria
- Real Estate Funds (preview): Real estate fund managers, their vehicles, and the bridge from fund capital to the properties, loans and lenders underneath. https://dfxintel.com/data-factory/real-estate-funds; standalone product https://dfxintel.com/real-estate-funds
- Private Credit (preview): Credit providers, BDCs and credit funds, the borrowers they finance, every facility as filed, and the sponsors behind the borrowers. https://dfxintel.com/data-factory/private-credit; standalone product https://dfxintel.com/private-credit
- Allocators (live): Who owns the capital: public and corporate pensions, Taft-Hartley plans, endowments, foundations, insurers and pools, their consultants, and the commitments they disclose. https://dfxintel.com/data-factory/allocators; standalone product https://dfxintel.com/allocators
- RIA Intelligence: US registered investment advisers (Form ADV since 2011), IAPD-registered advisors with dated registration history, private funds, advisor moves and teams, RIA successions and absorptions, state statistics. Canonical pages under https://dfxintel.com/ria (firms by CRD, advisors, funds, teams); the Data Factory domain page is https://dfxintel.com/data-factory/ria. Facts are labelled reported, fact from registration, or derived; no book size, no mobility prediction, nothing sold.
- Allocator Intelligence: the capital-owner graph behind private markets. Public pensions (every Census ASPP unit), corporate and Taft-Hartley defined benefit plans (Form 5500 with Schedules SB, H, C, D), endowments and foundations (IRS), state pools; reported assets, board-approved allocation targets and actuals (ACFR, 10-K XBRL), consultants and OCIOs (Form ADV, Schedule C, ACFR), investment staff, managers and fund commitments resolved to the PE and VC graphs, re-ups and first-time managers in the plans' own words, dated capital-flow events. Canonical pages under https://dfxintel.com/allocators (allocators, consultants, managers, funds, commitments, policy, activity). A target is never an actual, a disclosed holding is never an approval, estimates are labelled, nothing predictive is published.
- Private Credit Intelligence: the capital structure layer behind private markets. Every business development company's schedule of investments read from Inline XBRL since the SEC's 2020 tagging rule (184 BDCs on the tape), every quarter end from Dec 31, 2020, carrying each BDC's own fair value, cost, principal, spread, PIK, maturity and unfunded commitment per position, rolled into borrower groups and facilities (borrower, kind and lien) with their BDC lender groups, the quarter-over-quarter change ledger, sponsor attributions with their basis, the sponsor by lender graph, the maturity wall where filers tagged a maturity, and the credit providers and Form ADV credit funds behind them. Canonical pages under https://dfxintel.com/private-credit (lenders, BDCs by CIK, borrowers, facilities, maturity wall, sponsor by lender, sectors, funds, changes, methodology); the Data Factory domain page is https://dfxintel.com/data-factory/private-credit and entity pages are https://dfxintel.com/data-factory/entity/pc/<id>. Every size is the sum of the pieces the tape can see, a lower bound and never a facility's commitment; a maturity exists only where it was tagged; a mark is a mark; a sponsor named by a portfolio stem or a Form D signer is an inference and is labelled one; non-accrual is read from each BDC's own schedule footnotes (placed and returned to accrual, by BDC and quarter) and never inferred from a mark; covenants, amendments and default are not on the graph.
- Real Estate Fund Intelligence: the manager layer between private capital and property. Every real estate fund vehicle sworn on Form ADV Schedule D 7.B(1) since 2011 (more than 1,600 advisers and more than 12,000 vehicles ever filed; the current counts are on the page) with gross asset value by year, owners, master and feeder structure, minimum investment, first and last report and quarantined readings; managers classified from their own quoted website text; owner entities, properties, recorded loans, lenders and the public pension commitments that name them, bound to a manager under a named evidence class, with only rules that passed a blind-label gate published. Canonical pages under https://dfxintel.com/real-estate-funds (managers, vehicles, holdings, activity, lenders, lps); the Data Factory domain page is https://dfxintel.com/data-factory/real-estate-funds and entity pages are https://dfxintel.com/data-factory/entity/ref/<id>. GAV is gross asset value on a filing date as the adviser swore it, never fund size and never dry powder; a feeder's GAV is excluded where the same filing carries its master; a manager whose summed GAV exceeds three times its own regulatory assets is quarantined and enters no total; a name in common is never a link; a distress flag on one property is never a fund judgement; the property side reads as a floor. The read plane is registered; its MCP tools are pending.

## Limits

- It is not a fit for work that needs a licensed professional to sign it. It drafts; it does not certify.
- Browser work depends on the sites you are signed into. If a site's session expires, that job stops and asks you rather than guessing.
- Banks, brokerages and payroll sites are blocked in the Chrome extension's own code, not by policy. It cannot be pointed at them.
- Outbound sending inherits the deliverability of the mailbox you connect. It does not fix a domain with no authentication records.
- It is not the cheapest option. If your requirement is one narrow repetitive task with a stable interface, a point tool or a script will cost less and this will be over-specified.
- Usage is metered on top of the plan fee, so a month of heavy work costs more than a quiet one.

## Reference

- Knowledge base: https://dfxintel.com/ai
- Canonical product record: https://dfxintel.com/ai/general
- Entity record as JSON: https://dfxintel.com/ai/entity.json
- Capabilities as JSON: https://dfxintel.com/ai/capabilities.json
- Pricing: https://dfxintel.com/pricing
- Real Estate MCP developer reference: https://dfxintel.com/ai/real-estate-mcp
- Real Estate MCP capability catalog as JSON: https://dfxintel.com/ai/real-estate-mcp/catalog.json