Does Lever Remember Between Runs? The Candidate Profile Persists. The Screening Work Does Not.
Does Lever Remember Between Runs? The Candidate Profile Persists. The Screening Work Does Not. Short answer: Lever keeps the candidate. It does not keep the work done around that candidate. A profile holds contact details, resumes, and a history that spans every role that person was considered for. A scheduled agent still starts each run without what the last run decided, unless that decision was written back into a field the next run knows to read. That gap is easy to miss because the profile
Does Lever Remember Between Runs? The Candidate Profile Persists. The Screening Work Does Not.
Short answer: Lever keeps the candidate. It does not keep the work done around that candidate. A profile holds contact details, resumes, and a history that spans every role that person was considered for. A scheduled agent still starts each run without what the last run decided, unless that decision was written back into a field the next run knows to read.
That gap is easy to miss because the profile looks complete. The missing part is not a resume or a stage. It is the reasoning that connected several profiles together on one morning and was gone by the next.
What the candidate profile actually keeps
The durable layer in Lever is the candidate profile. It holds name, location, email addresses, phone numbers, and profile links, gathered from every job that person has been considered for at that company. Those details travel with the person to each new role, so a second application does not start from a blank contact card.
Resumes live in the same place. A resume attached to an application is stored automatically in the resume section of the profile, and several documents can sit there at once: the original submission, an updated version, a cover letter, a portfolio. The parser adds name, organization, and contact details from the file into the candidate summary, and the parsed text is what recruiter search runs against.
That persistence is real and useful. A recruiter opening a profile months later sees a person with a history, not a single isolated form. A scheduled step can read the same profile again tomorrow, because none of that data left the system of record.
Notice what kind of memory that is. It stores what was submitted and what was entered. It does not, on its own, keep a running account of what an automation concluded while working across many profiles at once.
What an opportunity keeps, and where it stops
Each application creates a separate record called an opportunity. The opportunity holds the posting it belongs to, the resume sent with it, the current stage, and the source that brought the person in. Beneath the profile sits a history of activity across all of that person's opportunities at the company, including interviews, feedback, and notes, so a reviewer can see how earlier applications went while judging a new one.
That structure answers a single-person question well: where does this candidate stand, and what happened last time? It answers a pipeline question poorly: what did last night's review learn about this group of candidates, and which of those lessons should change what happens first today?
Lever also works as a recruiting CRM, not only as a tracker for open roles. Profiles stay in the talent pool after a role closes, and recruiters can search them again for a future opening. The pool remembers people. It still does not remember the screening pass that sorted them: which near-duplicates were merged in judgment but not in the system, which source kept sending strong applicants for one team, or which three people were waiting only because feedback had not arrived.
Each record remembers the object it was built around. None remembers the work that moves between records, runs, and reviewers.
What a scheduled run does not keep
Picture a common setup. A scheduled agent runs early each morning over new applications for several open roles. It checks for duplicate profiles, flags people who applied to two roles at once, notes where interview feedback points in opposite directions, and prepares a short list of cases that need a human call.
On the first morning, that sorting is the work product. The agent learns that two profiles are the same person under two email addresses, that one posting draws a steady stream of well-matched applicants from one source, and that a small group is stalled for one shared reason rather than five separate ones.
When the run ends, the profiles are still there. The opportunities, stages, and feedback can all be read again tomorrow. The sorting, the source pattern, and the reason those cases were grouped together are not stored as a shared history the next run can load. Unless someone hand-builds a place for that output and teaches every run to write it and read it back in exactly the same shape, the second morning re-derives the groups, re-reads the feedback, and rebuilds the short list from the raw records again.
Repeated across roles and weeks, the cost is repeated reasoning. The record stayed perfect while the learning from working that record evaporated between runs.
Give the runs one shared memory
Vilix AI is built for that missing layer: one memory that sits outside any single profile or opportunity, and that every run can load before it starts and update before it ends.
It is cloud-hosted, so there is zero infrastructure to run or maintain. No side database has to be provisioned, keyed to candidate IDs by hand, and kept in sync just so an agent can keep its own screening notes.
Every connected tool reaches the same memory over MCP. A screening agent beside Lever, a drafting assistant preparing outreach, and a reporting step summarizing the pipeline can all read and write the same Vilix AI memory. The same memory follows the work everywhere, across tools and devices, instead of stopping at the boundary of one candidate profile.
Vilix AI stores full conversation history, not just facts pulled out of it. For hiring work, that distinction is practical. A single extracted fact can say a candidate was flagged. The full history shows what the run saw, what it grouped together, and why it asked for a human decision, so the next run can check the reasoning instead of repeating it.
With that layer in place, the morning routine changes. The first run saves its groups, exceptions, and open decisions to Vilix AI. The second run loads that saved work first, then spends its time only on new applications and on the cases still waiting. Lever keeps the record of record. Vilix AI keeps what the runs learned from working that record.
Start free, keep control of the data
Vilix AI has a free plan forever, with no time limit forcing a choice. Paid plans start with a 7-day Pro trial with no credit card required, so a full hiring cycle can be tested before any payment.
Data stays under the team's control. Everything can be exported at any time, and everything can be deleted at any time.
Lever will keep doing its part well. Profiles will hold their resumes and histories, and opportunities will hold their stages. Add one shared memory for the scheduled work around those records, and the reasoning survives too. The next run then starts where the last run stopped, instead of reading the same profiles and calling it a fresh start.