Services / Automation
We turn applications into one structured record, shortlist against criteria you own with the evidence attached, get interviews scheduled and outcomes sent on time, and raise onboarding work on signature. People still make every hiring decision.
Criteria owned by people. Evidence shown. No automated rejections.
The business case
Applications sit unread because reading them is slow, interviews take three weeks to arrange, and people who were never told anything stop answering. We remove the work that causes the delay, and keep the judgement where it belongs.
Steady applicant volume, screening that varies by who reads it, interviews arranged by hand across calendars, candidates left without an answer, or onboarding that starts after day one. Measured baselines exist or can be produced.
If you hire a handful of people a year or mostly through agencies, change the process by hand first. If your ATS is being replaced, decide that before automating on top of it.
Where this applies
These are example situations, not pre-packaged promises or claims about past results.
Parse CVs from portals, email and referrals into one structured record with role, experience, location, notice period and source, so screening starts from the same shape every time.
Learn: how many applications land per role, and how long before anyone replies?
Match each application against the criteria you set for the role, and show the evidence behind the match rather than a score with no reason attached.
Learn: what are your actual must-have criteria, and who owns them?
Coordinate availability, send reminders, handle reschedules and collect panel notes into one structured record per candidate.
Learn: how many days pass between a good candidate saying yes and the first conversation?
Send a clear outcome on schedule and, where you choose, considered feedback, so the talent pool stays warm and your reputation holds.
Learn: what share of applicants hear nothing at all?
Raise accounts, equipment, access and training tasks on signature, tracked to completion with a named owner for each item.
Learn: how long after signing can a new starter actually do their job?
Answer policy and process questions from your own handbook and route the cases that need judgement to a named person with the context attached.
Learn: how much of the people team's week goes to repeat questions?
How the work flows
Applications from every source are collected and parsed into structured fields, checked against the criteria written for the role. From there the shortlist passes a human review point: one path leads to an offer and an onboarding checklist with owners, the other to a clear outcome sent to the candidate.
Written criteria per role, evidence behind every match, a person deciding every outcome, and a runbook with named owners.
Automated rejections, scores nobody can explain, messages sent without approval, and screening built on records nobody maintains.
Groundwork
Before anything touches a candidate, we need to know how hiring really runs here: who decides, what the ATS actually holds, and what the people team is allowed to promise.
Where applications come from, where they stall, where candidates disappear and how long each stage really takes, traced through the roles you actually run.
What the ATS and HRIS hold, what lives in inboxes and spreadsheets instead, and which calendars, payroll, IT and finance systems onboarding depends on.
The criteria that define each role, who owns them, what must always be decided by a person, and how decisions are recorded and reviewed.
Time-to-hire, stage conversion, interview no-shows, offer acceptance, candidate response times and onboarding completion, measured now so any change can be checked against them.
Applications in
The point is not more data. It is that every application can be read, compared and answered without someone re-keying it.
Portals, career pages, email, referrals and event lists written into one place with the source and full history preserved.
Experience, skills, qualifications, location, notice period and salary expectations extracted from the document, with anything unclear left visible rather than guessed.
Existing candidates matched before new records are created, so previous applications and conversations stay with the person rather than being recreated.
Candidate consent recorded and expiry tracked to the retention period you set, with deletion carried through the systems that hold the data.
Matching & shortlisting
The system orders the work and shows its reasoning. Nobody's application is decided by a number nobody can explain.
Must-haves, nice-to-haves and disqualifiers written down with the hiring manager before screening starts, and versioned against the role.
Each match points at the line in the document that supports it, so a recruiter can check it in seconds and disagree with a reason.
Applications are ordered so the team works the strongest first, with anything ambiguous reaching a person rather than being filtered out.
The shortlist is agreed by people before anyone is written to, and the decision is recorded against the criteria used.
Interviews & decisions
The aim is fewer lost days between 'yes' and the first conversation, and a decision record that stands up afterwards.
Panel and candidate calendars coordinated, invitations and materials sent, and reschedules handled without anyone chasing three people.
Confirmed attendance, reschedule requests and no-shows tracked, with reminders timed to actually reduce the no-show rate.
Panel feedback collected against the same questions for every candidate, so the comparison is about evidence rather than who spoke loudest.
Outcome, rationale and criteria captured on the role, so a later question about a hire can be answered from the record instead of memory.
Outcomes, onboarding & the people team
Everyone who applied gets an answer, everyone who joins gets a working first week, and the people team gets its time back.
Rejections and offers sent on schedule, with considered feedback where you choose to give it, drafted from the panel's notes for a person to approve.
Strong candidates who were not hired kept with consent, and surfaced when a matching role opens rather than lost in a spreadsheet.
Accounts, equipment, access, payroll setup and training raised as tasks on signature, each with an owner and a due date, tracked to completion.
Policy and process questions answered from your own handbook, with anything sensitive, unusual or legally significant routed to a named person.
Tools & platforms
These are the platforms we typically assess, not a list you must adopt. Where a platform's API blocks a design, we say so before anything is promised.
We work inside the ATS you run where its API allows, and say plainly where a platform blocks a design.
Availability, invitations, reminders and outcomes handled through the channels your candidates already use.
Extraction and matching with the evidence shown, on models chosen for your data-handling rules rather than for demos.
Employee records, start dates and org structure joined to hiring and onboarding data on one set of definitions.
Access, equipment and account requests raised into the tools IT already works from, with owners and due dates.
Routing, approvals, reminders and retries matched to your hiring volume, hosting needs and what your team already runs.
Funnel, outcome and onboarding metrics with a single definition each, plus a record of who decided what and why.
Delivery approach
New flows run next to your existing process on live roles, with every match and every message reviewed by the people team before it counts. Nothing is switched over on a demo.
Agree the roles in scope, who owns the criteria, what must always be decided by a person, the data and retention rules, and the baseline figures to measure against.
Output: Scope note, criteria ownership and baseline measures
Walk the funnel with the people team and hiring managers, review ATS and HRIS data quality, and trace onboarding from signature to a productive first week.
Output: Funnel map, data quality findings and opportunity list
Decide what is parsed, matched, scheduled and sent automatically, where a person reviews and decides, and how outcomes and onboarding tasks are owned.
Output: Design note with criteria, review points and message rules
Build intake, matching, scheduling and onboarding flows, then run them alongside your current process on live roles, with every match and message reviewed.
Output: Working flows, parallel-run results and corrected matching rules
Phase the rollout by role family, train recruiters and hiring managers on the review points, and hand over the runbook, dashboards and ownership.
Output: Live flows, trained team, runbook and reporting pack
A people-team owner who can settle criteria and message rules, hiring managers willing to review shortlists, and access to the ATS, calendars and the systems onboarding depends on.
Scope, access and the rules for what must always be decided by a person are agreed in writing before build. Every criteria set and message rule is written down so you can change it and see what moves.
What you receive
Everything is delivered so your people team can change it without us, and so a hiring decision can be explained from the record.
Applications from every source in one shape, de-duplicated against your history, with consent and retention tracked.
Candidates ordered against criteria you own, each match pointing at the line in the document that supports it.
Interviews arranged and reminded without chasing, and every candidate given an outcome on schedule.
Accounts, equipment, access and training raised on signature with named owners and due dates, tracked to completion.
How each flow works, how to change criteria for a role, how to recover a failed run and who owns what.
We are not a recruitment agency and we do not take placement fees. The systems are built to your process; your people team stays accountable for who you hire.
Related services
These are the services people-team work usually connects to.
Questions before starting
No. Rejection is a human decision here by design. The system parses, orders and evidences applications, and prepares drafts for a person to approve. Nothing adverse is sent without someone deciding.
Criteria are written down and owned before screening, every match shows the evidence behind it, and decisions are recorded. We also review outcomes for patterns over time. Local employment law is still your counsel's call, and we will not claim to settle it.
Greenhouse, Lever, Workday, BambooHR, Zoho Recruit and Ashby where the API allows, plus exports where it does not. If a platform genuinely blocks the design, we say so before you commit.
Access is minimised and approved, retention follows the period you set, consent is recorded and deletions carried through, and nothing goes to a third-party AI service without your approval. Self-hosted models are an option where that is a requirement.
It removes the two things that usually lose them: waiting and silence. Faster scheduling and a clear outcome on schedule are what we measure, not a vague improvement in experience.
Intake, parsing and a criteria-based shortlist queue often run in production within weeks. Onboarding and offer flows take longer because they depend on IT, payroll and access approvals.
Yes, that is part of the job: duplicates merged, missing stages and owners filled, and criteria attached to live roles. We will not build screening on records nobody trusts.
Then automation probably is not the answer yet. Tell us your volume and we will say honestly whether a process change by hand is the better spend.
Start with one flow
Share your channels, your ATS and your last few roles. We will propose a scope, the criteria to agree and a first flow to put into production.