Sourcing bench for each open role, straight into Greenhouse

Build me a sourcing bench my recruiters open per open role, so sourcing a requisition happens on one screen instead of across twenty LinkedIn tabs and a spreadsheet nobody trusts. Crustdata is where candidates get discovered and enriched, Greenhouse is the system of record I push people back into, and the app itself owns the shortlists, fit notes, ratings and push history.

The app opens on a role picker. Pull my live requisitions with the Greenhouse List Jobs operation filtered to open jobs, and show each one with its title, department and office. Picking a role opens that role's bench, and everything from that point on is scoped to that requisition. Seed the search from the role itself: prefill the title from the job title and the location from the job's office, then let me correct or add to anything before I run a search. I should be able to move between roles without losing work I have already done on another one.

The search panel builds a candidate query out of title, skills, location and current company. Every one of those pickers is a type ahead backed by the Crustdata Autocomplete Person Search Fields operation, which is free, so I am always choosing values the dataset actually recognizes rather than typing keywords that quietly match nothing. Running the search calls the Crustdata Search People (In-DB) operation. Be explicit with me about cost, because that search bills per result returned while the autocomplete does not: let me set how many results I want before I run, show me the total size of the matching pool that comes back with the first page so I can judge whether to narrow the filters before paging deeper, and never silently fetch more pages than I asked for. Note that skills are a filter only field in that dataset, so I can search on them but should not expect them echoed back on the results.

Results render as cards, not a dense table. Each card shows the person's name, current title and company, location, and their career history built from the experience data already in the search payload, with previous companies, titles and date ranges so I can read a trajectory at a glance. Show me enough history to judge someone without opening anything else, and link out to their profile.

On engineering roles, each card gets an Add developer detail action. That calls the Crustdata Enrich Developer Profiles operation for that person and renders repositories, follower counts and linked handles beside the profile. Keep it as an explicit per person action rather than something the app runs across every result automatically, because it is an extra lookup per person and it is only interesting on technical roles. Once pulled, keep it on the card so I never fetch the same person twice.

From a card I move someone into this role's shortlist, which is where the real work happens. A shortlisted person carries a fit note in my own words and a rating. Shortlists are per recruiter: my shortlist for a role is mine, and another recruiter working the same requisition keeps their own alongside it. Let me sort and filter the shortlist by rating, and drop someone off it when I change my mind.

Contact details stay hidden until someone is shortlisted. Only then does a Reveal contact action appear, calling the Crustdata Enrich Person Contact Info operation to surface business email, personal email and phone numbers. This ordering is the whole point rather than a detail: contact lookups are the expensive part and I want them spent on the handful of people I have actually committed to, never on everyone I browsed past. Cache what comes back onto the shortlist entry so a second look does not cost a second lookup, and when no contact data is found say so plainly instead of showing an empty field that reads like a bug.

An Add to Greenhouse action on a shortlisted card writes the person back with the Greenhouse Create Candidate or Prospect operation. Default to creating them as a prospect, which is the sourcing case, and let me switch to creating a candidate with an application against this requisition when that is genuinely what I want, because Greenhouse treats those as two different things. Carry over their name, the contact details I revealed, their current company and title, and my fit note, so whoever picks them up in Greenhouse can see why they were sourced. Greenhouse attributes every write to a real user, so let me pick which recruiter gets credited from the Greenhouse List Users operation and use them as the acting user on the write.

Two rules have to be enforced by the app rather than left to recruiters remembering. First, stamp every pushed person with the date they were sent and the role they were attached to, and show that on the card. Second, once someone has been pushed for a requisition, grey them out everywhere on that role's bench, in fresh search results as well as on shortlists, for every recruiter and not only the one who pushed them, and block a second push with a message naming who sent them and when. Two recruiters working the same open role must not be able to double add the same person. That push history is shared across the team and has to persist.

On practicalities: Crustdata rate limits these endpoints separately and the enrichment calls are tighter than search, so pace per person lookups sequentially rather than firing them all at once when I open a busy shortlist, and show progress instead of appearing to hang. Handle a search that returns nothing with an empty state that suggests which filter to loosen. Never invent an email address, a phone number or a repository. If the lookup came back without it, show it as not found.