Candidate pipeline and ATS upkeep
An ATS is only worth searching if it is true, and keeping it true is work squeezed in between calls. An AI agent can take on that data entry, checking each batch of applicants against the records you already hold, writing the new ones and attaching the CV, so a consultant opens a job and sees everyone who applied.
The role in brief
Takes each applicant file, checks every person against the ATS, creates or updates the record, attaches the CV and reports what it could not place.
Your ATS, such as Loxo, and Microsoft Teams, where your people drop each applicant file and read what was done.
The consultant who owns the job. Possible duplicates, applicants with no CV and files that match no open job go to them, with anything a candidate needs to hear.
Applicants logged from each file, records held for a consultant, and candidates waiting at a stage.
ATS data entry falls to whoever is free
A resourcer does it, or a recruitment administrator, or the consultant at the end of the day. They export applicants from LinkedIn and the job boards, search the ATS for each name, key in the new ones, attach the CV and move the stage. Fitted in around calls, it is done late or in part, and the database drifts away from what is really happening on the desk.
Duplicates and half-filled records
The same candidate is entered twice, once from an advert and once from a referral, with the notes split between the two. A search for a skill misses someone you spoke to last year because their record has no CV attached. The agency pays for adverts to find people it already holds.
Applicants who are never logged
An applicant who is not in the ATS is not followed up and cannot be found for the next job. The strong ones take other offers while the list sits in a download folder, and the consultant who took the brief is working from memory and an inbox.
Records you have to stand behind
The Conduct Regulations require an agency or employment business in Great Britain to keep records showing it complies, for at least a year after it last acted for the candidate. Data protection law says candidate information must not be kept longer than you need it. Both are hard to show from records keyed in late.
How applicants get from a file to a true record
The AI worker logs applicants in your ATS, reports in Microsoft Teams and follows your own pipeline stages, mapped with you before the first file. Consultants keep every decision about a candidate and every conversation with one.
Intake
Takes the applicant file a person exports from LinkedIn and drops into Teams, reads every row and works out which open job in the ATS it belongs to.
A file it cannot match to one open job.
Duplicate check
Searches the ATS for each applicant before writing anything, so a returning candidate has their existing record updated and is not entered a second time.
A possible match it cannot settle. The consultant picks the right record.
Record
Creates or updates the candidate, attaches the CV where one was sent, adds them to the job's pipeline and tells the consultant in Teams what was done.
Nothing, where you have given a standing instruction for candidate records. Otherwise a consultant confirms each write.
Stage upkeep
Lists candidates who have sat at one stage longer than the days you set, drafts the update each is owed and answers consultants' questions about the records in Teams.
Every message to a candidate, and any stage change that records a decision about them.
Applicants logged from each file
Of the applicants in each file, how many were written to the ATS as new records and how many matched a record you already held. Reviewed with you each month.
Records held for a consultant
How many applicants the agent could not place on its own, such as a possible duplicate or a missing CV, and how long each waited for a decision.
Candidates waiting at a stage
How many candidates on open jobs have not moved stage for longer than the number of days you set, shown by consultant.
Questions owners ask
Does the agent log in to our LinkedIn account?
No. LinkedIn does not permit software that automates activity on its site, and we do not automate client LinkedIn accounts. One of your people exports the applicant list and drops the file into Teams. Everything after that is the agent's, from reading the rows to writing the records in your ATS.
Can AI log our applicants into Loxo or Bullhorn?
It can, using the API your ATS's maker publishes: Loxo, Bullhorn, Vincere and JobAdder each have one. Which route your licence opens is settled when the role is scoped; for any other ATS, the first question is whether its maker offers a way in on your licence.
Can it keep our ATS up to date without a consultant approving each record?
That is your choice. By default the agent proposes a write and a consultant confirms or cancels it in Teams. You can give a standing instruction that lets routine writes, such as new candidate records, go straight in. A possible duplicate always goes to a person.
Will the agent send messages to our candidates?
Not on its own account. It can draft the update a candidate is owed, and a consultant approves it before it goes. Phone calls stay with your consultants, as does every conversation about a role, and nothing is ever sent through LinkedIn. How long a candidate's record is kept is also your decision.