Integrations

Your stack. Any model. One score.

ScoringFactory plugs into the tools you already run. It reads the real record from your ATS, CRM, and data providers, scores every person and company against your bar on the model you trust, and writes the result, with receipts, back where your team works.

Your systems
ATS & CRM
GHGreenhouseCandidates and reqs
SFSalesforceAccounts and contacts
HSHubSpotDeal flow and ICP
Data
LILinkedInThe real record
APApolloEnrichment
CBCrunchbaseFirmographics
ScoringFactoryThe scoring engine
Score vs your bar
92/ 100
Your barRanked, with receipts
Runs on any model: OpenAI, Claude, Gemini, or your own

One score for every person and company, tuned to your bar. Each point traces back to the record that earned it.

Back to your tools
Output
GHGreenhouseScores on profiles
HSHubSpotFit as properties
SLSlackShortlist alerts
GSSheetsThe ranked list
How it fits in

Connect your systems. Run it on any model.

ScoringFactory does not replace your stack. It reads from the tools you already run, scores on the model you choose, and writes the result back where your team works.

Bring your systems

It reads the record you already have

Point ScoringFactory at your ATS, CRM, and data providers. It builds a sourced profile for every person and company, then writes the score and receipts straight back onto the record.

  • Candidates and open reqs from Greenhouse, Ashby, Lever, and more
  • Accounts, contacts, and deal flow from Salesforce and HubSpot
  • The real record from LinkedIn, Apollo, Crunchbase, and enrichment
  • Scores and evidence written back so nobody re-keys a thing
Use any model

Model-agnostic by design

The reasoning layer is yours to choose. Run scoring on a frontier model, in your own cloud tenant, or on a self-hosted open model. Your data is never used to train shared models.

  • OpenAI, Anthropic Claude, or Google Gemini out of the box
  • Azure OpenAI and AWS Bedrock to keep it in your own account
  • Open-weight models like Llama and Mistral for full control
  • Bring your own key, so nothing about the model is locked in
The full stack

Every tool we connect to

Sources feed the score. Destinations receive it. Bi-directional connections do both. Model is the reasoning layer you choose. Roadmap items are on the way.

Bring your whole stack. Get one score.

Connect the tools you already pay for, run scoring on the model you trust, and get a ranked, evidence-backed score written back where your team already works.