ICP prospecting workbench for building target account lists

I want an ICP account prospecting workbench I can open whenever I need to build a target account list, instead of exporting CSVs from a data provider and hand cleaning them in a spreadsheet. It is an internal tool for me and my sales team, not something that runs on a schedule. I drive the search, review what comes back, and choose what lands in the CRM.

The main screen is a filter panel next to a results table. The filter panel has industry, employee count range (a minimum and a maximum), HQ country, HQ city, founded year, and a free-text keyword field. When I hit Search, a handler calls Coresignal Search Companies (Filter) on the base company dataset with whatever fields I filled in, leaving blank fields out of the query entirely. That search returns matching company IDs plus total-count and paging headers, so show me how many companies matched and let me narrow the filters before pulling any full records.

The results table shows company name, employee size, headcount growth, industry, and website. Coresignal search returns only IDs, and every full record collected costs one credit, so only collect the rows actually on screen. For each ID on the current page, call Collect Company by ID on the multi-source dataset and fill in that row. Page the results using items_per_page and the x-next-page-after cursor rather than pulling everything at once, and cache collected company records so paging back and forth, or reopening a saved search, never spends credits on the same company twice. Prefer the multi-source dataset variant for anything shown to me, since it is the most current.

Clicking a row opens a detail drawer with the full firmographic record for that company: description, founded year, HQ location, employee size, industry, website, and the headcount trend over time. Below that, show the decision makers at that company. Fetch them with Coresignal Search Employees (Filter) scoped to that company, then Collect Employee by ID on the multi-source dataset for each one, so I can see name, title, seniority, and tenure in role. Default the employee search to senior titles, director level and above, with a control to widen it. Each decision maker has a checkbox.

I select companies in the table with checkboxes and click Push to HubSpot. For each selected company, first run HubSpot Search Companies filtered on the website domain so I do not create duplicates. If the company already exists, reuse that record and report it as matched rather than created. If it does not exist, use HubSpot Create Company with name, domain, industry, employee count, city, and country. For every decision maker I ticked in the drawer, use HubSpot Create Contact, or Batch Create Contacts when I am pushing several at once, then use Create Association to link each contact to its company. When the push finishes, show a summary of what was created, what was matched to an existing record, and anything that failed, with a retry option for the failures.

Let me save a named ICP search, for example "Series B fintech, 50 to 200 employees, UK", so I can rerun last month's criteria in one click. Saved searches appear in a list I can open, rename, and delete. They store the filter values rather than the results, so every rerun returns fresh data. Saved searches are per user.

Persist an "already pushed" marker per company, keyed on the company website domain, and store the HubSpot record id alongside it. Any company already in the CRM renders greyed out in the results table with a small "in HubSpot" label, has its checkbox off by default, and offers a link to open the HubSpot record directly. This marker persists across sessions and is shared across the whole team, so two reps do not push the same account twice.

Throughout, treat Coresignal credits as a scarce resource worth protecting: never collect a full record the user has not asked to see, surface the credits-remaining value returned on responses somewhere unobtrusive in the UI, and handle an insufficient-credits response with a clear message rather than a silent empty table.