Train title-normalisation and company-matching models on structured profiles
Give an AI or data-tools team labelled examples to train models that normalise job titles into function and seniority and match messy company names to one company.
- Free API key
- No credit card
- Run your sample
What it costs: Charged per request: credit pricing.
- People data feed
- Company data feed
- Data feed
- Feed licensee
- AI agents & developer tools
How the data flows
- labelled title examples
- VP Revenue Operations
- name
- Jane Example
- mobile
- +1 (415) 555-0142
- work email
- jane@example.com
- company
- Example Inc.
Steps
- Load
people_feedroles (role_title,organization_name,forager_organization_id) andcompany_feed(name, domain,linkedin_slug, industry) - Build title examples labelled with the function and seniority mapping the team defines
- Build company-name pairs that resolve to the same
forager_idfor entity-resolution training - Hold out a test set from a later delivery and track accuracy on each refresh
Inputs and outputs
| Inputs | Outputs |
|---|---|
| people feed | labelled title examples |
| company feed | company name match pairs |
| label taxonomy | evaluation set |
Questions about this recipe
Where can I get training data for a job title classifier?
This recipe uses Forager People data feed and Company data feed. People data feed: A licensed file of 800M+ professional profiles, refreshed on a schedule and delivered to Snowflake or S3. Give an AI or data-tools team labelled examples to train models that normalise job titles into function and seniority and match messy company names to one company.
What does it cost to train title-normalisation and company-matching models on structured profiles?
People data feed is a flat annual licence, global or regional, not priced in credits. Self-serve plans are Starter, Growth, Pro and Scale. See plans and prices
What do I need to run this recipe, and what comes back?
You send company feed, label taxonomy and people feed. You get back company name match pairs, evaluation set and labelled title examples, as JSON. You also need a Forager API key, which comes with a free account.
Which dataset helps train company name matching?
This recipe uses Forager People data feed and Company data feed. People data feed: A licensed file of 800M+ professional profiles, refreshed on a schedule and delivered to Snowflake or S3. Give an AI or data-tools team labelled examples to train models that normalise job titles into function and seniority and match messy company names to one company.
Is there a free way to try Forager on this workflow?
Yes. Request a sample extract of the people data feed on a 30-minute call and check the fields against your own records before you license it.
When should I run this workflow?
When a team builds or retrains a title classifier or an entity-resolution model. Models learn from real-world title and company variation instead of a hand-built sample.
Which Forager products does this recipe use?
People data feed and company data feed. People data feed: a licensed file of 800M+ professional profiles, refreshed on a schedule and delivered to Snowflake or S3. Company data feed: a licensed file of 65M+ company profiles and firmographics, delivered to Snowflake or S3 on a schedule.
How accurate is Forager data?
The test that counts is your own: send a list of profiles you care about. We run it and show you what we find, number by number, so you can measure hit rate and connect rate yourself.
Which tools can run this workflow?
Delivered as a data share to Snowflake, Databricks and Amazon S3. Or call the endpoints in the steps above from your own code with one API key.
How many steps is this workflow?
4 steps: load people_feed roles (role_title, organization_name, forager_organization_id) and company_feed (name, domain, linkedin_slug, industry), build title examples labelled with the function and seniority mapping the team defines, build company-name pairs that resolve to the same forager_id for entity-resolution training and hold out a test set from a later delivery and track accuracy on each refresh.
Is there a contract or minimum commitment?
Data feeds are licensed on an annual agreement, with the term, volume and refresh schedule written into it. There are no self-serve plans for feeds.
Can we show Forager data inside our product?
Yes, under platform terms. Self-serve plans cover internal use by your own team. Showing Forager data to your users, or passing records on, is licensed separately: talk to sales about volume and which fields your users need to see.
Written answers are not available right now. The matching pages above still help, or book a demo and ask the team.
- $5 free credits
- No credit card
- Run your sample
More recipes for LLM apps
-
License a people feed for retrieval over people and company records
-
Ground AI CRM agents with fresh company and people records
-
Brief users on the people they are about to meet, from an email address
-
Refresh contact data before an agent writes to the CRM
-
Resolve a sign-up email to a person inside a copilot
-
Answer hiring and labor-market questions in an AI analyst from a job postings feed