7 updates
· 7 updates
AI job descriptions, every pipeline in one table, and transcripts on every spoken answer
Draft a job description from the title, run your pipelines from a proper table, put assessment questions in the order you meant, and read the transcript behind every spoken score.
PipelinesAIAssessmentsAnalyticsIntegrationsATS
Hivemind runs your hiring end to end: pipelines, assessments, outreach, interviews and offers in one place. This release is mostly about the parts you touch every day.
Creating a pipeline asks for less typing and drafts the job description for you. The pipelines page became a table you can actually work in. Assessments run in the order you set, and every spoken answer now shows the transcript behind its score. Dashboards are yours to start, and a connected AI assistant can answer across your whole company. At the end, for teams running Ashby, there is a two way sync.
This page walks through each feature: the problem it replaces, how it works, and where to find it in the app. Everything here is live in Hivemind today.
Write the title, get the job description
Every pipeline needs a job description, and writing one from an empty box is the step that keeps a role sitting in drafts for a week.
Now the create form asks for the job title, seniority, location and the rest in one place, and Draft with AI writes the description from it. You get markdown you can edit, not a locked block of text, and Preview posting shows exactly what a candidate will read before anyone applies.
The same fields and the same drafting button are now in an existing pipeline's settings, so a role you wrote months ago can be brought up to date without rebuilding it.
Where to find it: New Pipeline, on the Manual tab, and an existing pipeline's General settings.
Every pipeline in one table
A grid of cards looks fine with four pipelines and becomes a scrolling problem at forty.
The pipelines page is now a table. Sort it, filter it, and see the numbers that tell you where to look: candidates in each pipeline, how many are in progress, how many were hired. Tabs split active pipelines from drafts and closed ones.
Every row carries what you would otherwise go hunting for: edit, the dashboard, applicants, the application link, activate or deactivate, duplicate, and share to LinkedIn on anything active. Columns you do not need are hidden by default and can be switched back on.
Where to find it: Pipelines in the sidebar.
Assessments in the order you meant
Questions used to appear in an order the system chose. That is fine for a quiz and wrong for an assessment that should open gently and build.
Now you drag questions into the order candidates will answer them, or move one with the row menu when dragging is awkward. The order you set is the order the candidate sees.
On the candidate's side, the progress panel became navigable. They can jump to any question, see what is answered, what is pending and what they marked to come back to. It stays put while a video answer is recording, so nobody loses a take by clicking away.
Where to find it: The assessment builder, and the progress panel on the candidate's assessment.
See what the grader actually heard
Spoken answers were the one place you had to take the score on faith. The candidate talked, a number came back, and the only way to check it was to watch the recording yourself.
Now every spoken or video answer carries the transcript the grader read, collapsed until you want it, next to the score and the per-part reasoning behind it. When a recording has no speech in it, it says so plainly and scores zero rather than inventing a grade from silence.
It is the difference between trusting a number and being able to check one.
Where to find it: The assessment report on any candidate, on spoken and video answers.
Dashboards you start yourself
Boards used to arrive one of two ways: the six that ship with every account, or one Hive built because you asked for it. There was no way to simply start one.
Now there is a New dashboard button. Begin from blank and ask Hive to fill it, or start from one of the built-in boards and change it into what you actually want. Boards you make start as drafts, and publishing one gives you a link you can share outside Hivemind.
Every board also takes a pipeline and a time range now. Narrow to one role, or to the last thirty days, and the widgets about candidates, interviews and assessments follow. Widgets about the pipelines themselves stay as they are, because narrowing those to one pipeline would tell you nothing.
Where to find it: Dashboards in the sidebar, then New dashboard.
Questions about your whole company, not one pipeline
Connecting an AI assistant to Hivemind was useful for one role at a time. Ask about applications across the company, or about a list you saved last month, and it had nothing to work with.
This release closes that. Your assistant can now read how many applications arrived between two dates across every pipeline, open a saved candidate list and tell you who is on it, rename a list, take people off it, and rank the candidates in a pipeline by whichever score you care about, saying plainly who has no score yet rather than quietly dropping them.
It can also do something with the result: add people already in Hivemind to an outbound campaign by naming a list, instead of you exporting a spreadsheet and pasting addresses back in.
The safety rules got stronger at the same time. Removing people from a list, deleting a list and saving a shortlist now ask you first, and the rules the assistant follows are sent to it on connection rather than relying on a file somebody remembered to install. Candidate search also got a lot quicker.
Where to find it: Any AI assistant already connected to Hivemind. Connections are added in the assistant itself, and reviewed or revoked under Settings → Apps & Integrations.
Ashby candidates, without the copying
Running two systems usually means running them twice. An applicant arrives in your ATS, somebody retypes them into the tool that does the screening, and the results live in a third place that nobody looks at when the decision gets made.
Connect Ashby once and that stops. Applicants and their resumes arrive in the Hivemind pipeline you route them to, and everything Hivemind produces goes back the other way: assessment scores, cognitive and personality results, resume scores and AI interview outcomes, each written onto the Ashby candidate as a note. Hires and archives in Ashby are mirrored onto the Hivemind record, so neither side drifts.
Two things are optional and off until you turn them on. Hivemind can move a candidate forward to the matching Ashby stage as they progress, and it only ever moves forward, never back. And for a brand new Ashby job, Hivemind can build the pipeline itself: draft one for you to review, or build it and switch it on. Read the warning on that second option before you pick it, because a live pipeline sends real email and runs real assessments from the moment it starts.
One thing to know: only Ashby is connected this way today.
- Open Settings → Apps & Integrations and connect Ashby with your API key.
- Add the webhooks Hivemind gives you inside Ashby, so candidate updates reach it.
- Choose which Ashby jobs feed each pipeline, and decide whether new jobs get a pipeline automatically.
Try it yourself
Everything on this page is live in Hivemind today. Open the app and try it on a real pipeline, or book a walkthrough with the team.