Explorium AgentSource
Sales IntelligenceExplorium AgentSource is a B2B data platform built for AI agents and go-to-market teams. Agents match and enrich companies and prospects across twenty datasets, search by firmographic filters, track business signal events and run AI research jobs with a defined output schema.
What This Integration Enables
Explorium AgentSource is data infrastructure that assumes a machine is the customer. Its search surface takes structured filters rather than a query box, its enrichment surface takes resolved identifiers rather than names, and its research surface takes an output schema so the answer comes back shaped for the step that consumes it. That last detail is the tell: a product built for a person to read would return prose. This one is built to be called.
FlowRunner agents use it in three modes that fit different points in a workflow. Matching and enrichment turn something you already know about a company or a person into the twenty datasets Explorium holds, from firmographics and technology stack through funding history, website traffic, workforce trends, and analysis derived from public filings. Discovery goes the other way, building an audience from filters, with a statistics call that returns counts and distributions without returning records. Signal events and enrollments turn the platform into a watcher, so a funding round or a contact changing employer becomes something a flow reacts to.
Every one of those modes spends a metered allowance, and this connector deliberately exposes the meter as well as the data. That is a design choice worth naming, because an agent that can spend without seeing the balance is an agent whose budget is a surprise.
Without FlowRunner
With FlowRunner
Use Case Scenarios
-
An account list that resolves before it is enriched
A list of company names arrives from a conference export in Google Sheets. The agent calls Match Businesses in batches, supplying a domain wherever the export has one because a domain matches far more reliably than a name alone, and keeps the returned identifiers. Only the resolved rows go on to Enrich Businesses for firmographics and technology stack. The unresolved rows are not guessed at and not enriched; they go to a review sheet with what was known about them. The enriched records write into HubSpot or Attio, and the count of rows that failed to match is reported as a data quality number rather than quietly dropped.
-
Sizing an audience before paying for it
A campaign brief describes a segment in words. The agent turns it into filters, using Autocomplete Filter Values with semantic search to find the taxonomy entries Explorium actually accepts rather than guessing at wording. It then calls Get Business Stats, which returns how many companies match and how the matches distribute across each filter dimension, without returning any companies. The team sees that one filter is doing almost all the narrowing and loosens it. Only when the shape looks right does the flow call Fetch Businesses, and it tunes in preview mode first, because full mode returns complete records and is billed as such and nothing in the response says which mode was asked for.
-
A signal that starts a sequence
Selected accounts are enrolled for event monitoring. When Explorium reports a funding round or a sharp change in engineering headcount, the flow wakes, pulls the account's current owner from the CRM, runs Research Businesses with a fixed output schema so the answer has the same keys every time, and drafts an opening message. The draft lands with the rep in Slack next to the signal that caused it. If the rep approves, the sequence starts in Instantly or lemlist. The signal is automatic. The decision to act on it is not.
Human-in-Loop Highlight
Explorium allows one webhook per partner account. Add Webhook does not add a second destination, it replaces the existing one, so two flows cannot each register their own. That turns an ordinary looking setup call into an account-wide change: an agent told to start monitoring a new set of accounts, doing the obvious thing, silently repoints every event delivery every other flow in the company depends on, and nothing errors. The matching hazard sits on the other side, because Delete Webhook does not cancel enrollments. Standing subscriptions carry an ongoing cost, so removing the destination leaves them running and billing with nowhere to deliver.
So the agent may enroll and unenroll nothing until the destination question is settled. It calls Get Webhook first, runs Check Webhook Connectivity against the URL it intends to use, and when a different destination is already registered it stops: "Event enrollment for 84 accounts is ready. This partner account already points its webhook at the lifecycle flow's endpoint. Registering mine replaces theirs, and their enrollments will keep billing with nowhere to deliver. Repoint, or route through the existing endpoint?" A person answers, because the question is about who owns a shared slot, and that is an ownership question rather than a data question. This is the shape of human-in-the-loop that a metered, single tenant resource asks for: not approval of the data, approval of the claim on the resource.
Agent Capabilities
31 actionsMatching
2- Match Businesses Resolves companies to Explorium business identifiers from whatever you know about them: name, domain, website URL, or profile URL. Every enrichment operation takes those identifiers, so this is the first call in almost any business flow, and supplying a domain matches far more reliably than a name alone.
- Match Prospects Resolves people to Explorium prospect identifiers from a name, email, phone number, or profile URL. Supplying the business identifier of the person's employer alongside their name matches far more reliably than a company name string, so match the company first where you can.
Business Enrichment
4- Enrich Businesses Enriches one or more companies with a chosen Explorium dataset on the current route. Seventeen datasets are available on the same call, from firmographics and technology stack through funding history, website traffic, workforce trends, and analysis derived from public filings. Supply the business identifiers that matching resolved.
- Bulk Enrich Businesses Enriches a batch of companies with one dataset through the bulk route, which returns every result in a single response rather than as a job. Use it when the batch is large enough that one call per company is wasteful but small enough that you would rather not poll.
- Start Business Enrichment Job Queues an asynchronous enrichment over a previously uploaded list of business identifiers and returns a job identifier. This is the route for large sets, and it does not accept identifiers inline: upload them first, then start the job with the list identifier that upload returns.
- Enrich With Custom Enrichment Runs one of your account's custom enrichments over a set of entity identifiers. A custom enrichment is configured inside Explorium rather than through this API, so the identifier for it comes from your account, and a bulk mode toggle switches the call to the batch route.
Prospect Enrichment
3- Enrich Prospects Enriches one or more people with a chosen Explorium dataset: contact information, full professional profiles, or their recent social posts. Supply the prospect identifiers that matching resolved.
- Bulk Enrich Prospects Enriches a batch of people with one dataset through the bulk route, returning every result in a single response rather than as a job.
- Start Prospect Enrichment Job Queues an asynchronous enrichment over a previously uploaded list of prospect identifiers and returns a job identifier. Upload the identifiers first, because this route takes the list identifier that upload returns rather than inline identifiers.
Discovery
5- Fetch Businesses Searches Explorium's company database by structured filters and returns matching companies. This is prospecting rather than lookup: filter by size, revenue, country, industry taxonomy, technology stack, website keywords, intent topics, and recent events. Preview mode is for tuning filters, because full mode returns complete records and is billed as such, and paging is by the cursor the response returns.
- Fetch Prospects Searches Explorium's people database by structured filters and returns matching prospects. Filter by job level, department, title, skills, interests, tenure, and by the attributes of the company they work at. Set the contact details filter when you only want people whose email or phone Explorium actually holds.
- Get Business Stats Returns how many companies match a set of filters and how they distribute across each filter dimension, without returning the companies themselves. Call it before fetching to size an audience and to see which filter is doing the narrowing, which is much cheaper than fetching records in order to count them.
- Get Prospect Stats Returns how many people match a set of filters and how they distribute across each dimension, without returning the prospects themselves. Use it to size an audience before spending the allowance on records.
- Autocomplete Filter Values Returns the valid values for a filter field, which is how you discover what a taxonomy filter will actually accept. Semantic search matches on meaning rather than on prefix, which finds the right entry when you know what you want but not how Explorium words it.
AI Research
2- Research Businesses Runs Explorium's research pipeline over a set of companies and returns the answer as structured data. Supply a query describing what you want to know and, ideally, an output schema pinning the shape of the answer, because without one the model chooses the shape and it can differ between runs, which breaks a downstream step expecting fixed keys.
- Research Prospects Runs Explorium's research pipeline over a set of people and returns the answer as structured data. As with company research, supply an output schema whenever a later step depends on the shape of the answer.
Signal Events
4- Fetch Business Events Returns recent signal events for a set of companies over a time window: funding rounds, public listings, new offices, partnerships, hiring and headcount movement by department, awards, outages, cost cutting, acquisitions, and legal issues. These are the events a go to market flow reacts to.
- Fetch Prospect Events Returns recent signal events for a set of people: a role change, a move to a new company, or a job start anniversary. A contact changing employer is the classic trigger for re-engaging someone you already know at their new company.
- Manage Business Event Enrollments Manages which companies Explorium watches for events and pushes to your webhook when one occurs. It lists the current enrollments, starts watching, changes which events are watched, and stops. An enrollment is a standing subscription rather than a query, so it carries an ongoing cost for as long as it exists, and the destination has to be configured before it is worth starting.
- Manage Prospect Event Enrollments Manages which people Explorium watches for role changes, company moves, and job anniversaries, pushing to your webhook when one occurs. It works the same way as the business form, and an enrollment is likewise a standing subscription with an ongoing cost.
Asynchronous Jobs
4- Upload Entity IDs Uploads a list of business or prospect identifiers and returns a list identifier. The asynchronous enrichment routes take that list identifier rather than inline identifiers, so this is always the first step of a large enrichment.
- List Jobs Returns the asynchronous enrichment jobs on your account with their current state. Use it to find a job identifier you did not keep, or to see what is still running before starting more work.
- Get Job Status Returns the state and progress of one asynchronous enrichment job, and its results once it has finished. Poll it after starting a job unless you supplied notification settings instead.
- Cancel Job Cancels a running asynchronous enrichment job. Work already completed is still billed, so canceling limits further spend rather than refunding what has been spent.
Credits
3- Get Credits Summary Returns the active credit balance on the account. Explorium meters every enrichment and every full mode fetch against this allowance and different datasets cost different amounts, so read it before a flow that fans out over a large list.
- Get Credit Aggregation Returns credit consumption broken down over a period, so you can see which datasets and which flows are spending the allowance, and find the enrichment that is quietly costing the most.
- Get Credits Menu Returns the credit price of each dataset for your account, which is what tells you the relative cost of one enrichment against another before you run it at scale.
Webhooks
4- Add Webhook Registers the URL Explorium pushes enrollment events to. There is one webhook per partner account, so calling this again replaces the existing destination rather than adding a second one, and two flows cannot each register their own. The payload format setting controls whether the body arrives as a JSON object or as a JSON string, which matters to whatever receives it.
- Get Webhook Returns the webhook currently registered for a partner account, with its URL, headers, and payload format. Read it before registering anything.
- Delete Webhook Removes the webhook registered for a partner account. Existing event enrollments are not canceled by this, so they keep accruing with nowhere to deliver, and removing them is a separate deliberate act.
- Check Webhook Connectivity Asks Explorium to send a test delivery to a webhook URL and report whether it succeeded. Run it before enrolling entities, because an enrollment that cannot deliver still consumes its subscription.
Frequently Asked Questions
What can FlowRunner do with Explorium AgentSource?
FlowRunner agents can run Enrich Businesses, Bulk Enrich Businesses, and Start Business Enrichment Job in Explorium AgentSource, plus 28 more actions.
Does connecting Explorium AgentSource to FlowRunner require OAuth?
No. Explorium AgentSource connects to FlowRunner with an API key, no OAuth flow required.
Can Explorium AgentSource trigger a FlowRunner workflow automatically?
Explorium AgentSource doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.
Other Sales Intelligence integrations
Start building with Explorium AgentSource
Free plan, no card required. Connect in minutes.