Home · Solutions · Other solutions

Solution · Other solutions

Panels found, interviewers briefed and scorecards in before a candidate hears from anyone else

Interviews booked and feedback collected the same day

When a candidate reaches the interview stage, the panel's availability, the invitations, the interviewer briefs and the scorecard chase run themselves; recruiters keep the judgement.

DepartmentalMicrosoft TeamsHuman in the loopDeterministic automation
520interviews a month are arranged by hand at this illustrative engineering services firm, panel by panel, in email threads that nobody owns.

Executive summary

Challenge

Nine emails to book one interview, and feedback that arrives after the candidate has accepted elsewhere.

What changes

The trigger is the one event recruiting already produces reliably: a candidate moving to the interview stage in the ATS. From that moment a robot…

Business value

Time from the interview stage to the offer shrinks, because finding an hour takes minutes of machine work instead of days of email.

Systems involved

the ATS (slots, scorecards, referrals); Exchange Online calendars; Microsoft Bookings

Business problem

Recruiting

Interview coordination is a scheduling problem wearing a recruiting problem's clothes. Recruiters spend their days proposing times, proposing them again when one interviewer declines, and chasing scorecards that interviewers fully intended to write. The judgement in hiring, who is worth meeting and who gets the offer, takes a fraction of the elapsed time; the rest is calendars.

Interviewers arrive without the CV and without the competency questions, so the first five minutes go on asking the candidate what the role involves. Feedback is written days later from memory, which makes the panel's comparison weaker than the interviews were. Referrals, the cheapest source of hires anyone has, arrive as personal emails to whoever the employee happens to know in HR and disappear there.

The problem lasts because the pieces sit in different systems and nobody owns the number that would expose it. The ATS holds candidates and stages but not calendars. The calendars hold availability but not the process. No report says how many days pass between the decision to interview and the offer, so the delay is discussed as a feeling about recruiting rather than as a measurement of coordination.

How it works today

The sequence below is what most professional-hire loops look like once a recruiter decides somebody is worth meeting.

  1. SystemA recruiter moves the candidate to the interview stage in the ATS, which is where the supported process ends
  2. PersonThe recruiter emails three interviewers asking for availability, starting a thread that will run for days
  3. WaitingTwo reply, one does not; the thread is chased on Thursday and again on Monday
  4. PersonTwo options go to the candidate, who takes one, and then an interviewer declines it, so the cycle restarts
  5. PersonInvitations are sent by hand, with the CV attached if somebody remembers to attach it
  6. WaitingAfter the interview the recruiter asks for feedback and waits; it arrives days later as free text written from memory
  7. PersonThe hiring manager decides, the recruiter updates the ATS, and the timestamps that would explain the delay stay inside inboxes
SystemPersonWaiting

Why the current process costs more than it appears

The bill that never reaches the budget.

  • Coordinator hours are the part anyone can count, and they are not the expensive part. The offer that lands after the candidate has accepted elsewhere costs a full search, and it never appears as a line in any budget.
  • Roles that stay open weeks longer are paid for by the delivery organisation in overtime, in slipped projects and eventually in an agency fee, which is the most expensive way a company can buy coordination.
  • Feedback written three days late is worse feedback. Panels compare recollections instead of observations, and the weakest candidate assessment in a loop is usually the one that was written last.
  • Referrals arrive as private favours rather than as tracked applications, so the referrer is never credited, the candidate is never answered, and the cheapest hiring channel quietly stops being used.
  • Hiring managers describe recruiting as slow when the slowness is coordination, but nobody can show that, because the evidence lives in mail threads rather than in a system.

Cost of inaction

Twelve months of coordination at today's interview volume≈ €55,910
Interviewer hours spent inside scheduling threads, same year≈ €44,928
Both clocks together, one hiring year≈ €100,838

Recruiting is the one internal process whose deadline is set by somebody else. A candidate who books a dentist online in thirty seconds reads a week of silence as a verdict on the company, and every candidate lost that way restarts a search at full cost. Nothing in the rows above prices that, and nothing in them prices the vacancy weeks the delivery organisation absorbs while a role stays open.

Agency use is where this eventually surfaces in the budget, as fees paid to fill roles the internal process was too slow to close. It arrives labelled as a sourcing problem, which is the wrong diagnosis, and it grows with every requisition the same two coordinators are asked to carry.

Illustrative scenario

A plausible organisation with realistic proportions. The figures are there to be recalculated on your data; they are not a client result.

Organisation

A technology and engineering services firm with 2,100 employees; about 380 hires a year handled by four recruiters and two coordinators, requisitions and candidates in Workday Recruiting, Microsoft 365 across the company.

Volume

520 interviews a month, mostly panels of two or three engineers, booked in Outlook by hand; feedback and referrals arrive by email, and the intranet's open-roles page is updated when someone remembers.

Current process

Availability is gathered by email, slots are offered to the candidate and withdrawn when an interviewer declines, invitations are assembled manually, and feedback is requested after the fact and chased.

Bottleneck

24 minutes of coordinator effort per interview, plus about six minutes of interviewer time per interview spent inside scheduling threads. The head of engineering has asked why a loop containing four hours of interviews takes three weeks of calendar.

Solution

The stage change in the ATS triggers a robot that reads panel availability through Microsoft Graph, offers the candidate the slots the rules allow, creates the invitations, briefs each interviewer in Teams the day before and collects the scorecard afterwards, with an SLA task when it does not arrive.

Potential outcome

In the modelled case scheduling stops consuming days, interviewers arrive with the CV and the questions in front of them, and feedback lands the same day. The figures behind that are a model, not a measurement.

Proposed solution

The trigger is the one event recruiting already produces reliably: a candidate moving to the interview stage in the ATS. From that moment a robot takes the coordination. It reads the panel from the requisition, reads free/busy across those calendars through Microsoft Graph, applies the rules recruiting agreed (no interviews before nine, none in the last hour of a Friday, two interviewers minimum, a buffer around the panel's own meetings), and publishes only the slots that survive those rules to a Microsoft Bookings page the candidate can use.

The candidate's choice does the rest. Outlook invitations with a Teams meeting are created for the panel and the candidate, and a reschedule repeats the same search rather than starting a new thread. The day before, each interviewer receives an Adaptive Card in Microsoft Teams carrying the CV, the role profile, the competency questions for that stage from the SharePoint question bank, and the scorecard link. After the interview the same card asks for the scorecard, which is a Microsoft Form, so it takes a minute rather than an evening.

Feedback that does not arrive within a day becomes a UiPath Action Center task with an SLA: a reminder to the interviewer first, then an escalation to the hiring manager. Scorecards are written back to the ATS as the record of the decision. Two smaller cards run on the same plumbing: one submits a referral into the ATS tagged to the referrer, and one shows live open roles so employees stop referring people to positions that closed in March. Power BI turns the whole thing into days per stage, feedback compliance and interviewer load.

Native capabilities used

Microsoft Graph calendar availability and Microsoft Bookings pages; Outlook invitations with Microsoft Teams meetings; Teams Adaptive Cards; Microsoft Forms; UiPath Action Center SLA tasks completed in Teams; UiPath Orchestrator triggers, retries and audit

What we build

The scheduling rules, the candidate slot flow, the interviewer brief and hiring-manager pipeline cards, the scorecard mapping into the ATS, the feedback SLA and escalation logic, the referral and open-roles cards, and the funnel reporting

Custom integration

UiPath Integration Service connectors for Workday (or SAP SuccessFactors Recruiting over OData), Microsoft Outlook 365 and Microsoft Teams; the SharePoint competency question bank

How the automated process works

  1. AutomationThe stage change in the ATS starts the process; a robot reads the panel from the requisition together with the scheduling rules
  2. SystemFree/busy across the whole panel is read through Microsoft Graph, and the candidate receives a Microsoft Bookings page holding only the permitted slots
  3. AutomationThe chosen slot creates Outlook invitations with a Teams meeting; a reschedule repeats the same search instead of opening a new thread
  4. AutomationThe day before, each interviewer gets an Adaptive Card in Teams with the CV, the role profile, the stage's competency questions and the scorecard link
  5. PersonThe interview takes place and the interviewer completes the Microsoft Forms scorecard from the card
  6. PersonWhere no scorecard arrives within a day, an Action Center task reminds the interviewer and then escalates to the hiring manager
  7. AutomationScorecards are written back to the ATS, and Power BI reports days per stage, feedback compliance and interviewer load
AutomationSystemPerson

Human-in-the-loop model

Automation handles

  • Finding availability across the whole panel within the rules recruiting agreed
  • Candidate self-scheduling, invitations, Teams meetings and reschedules
  • Interviewer briefs with CV, role profile, competency questions and scorecard link
  • Feedback collection, reminders and escalation to the hiring manager

People decide

  • Who sits on the panel and how the loop is designed
  • The interview itself and everything written on the scorecard
  • The hiring decision, the offer and its terms
  • Exceptions: executive searches, candidates who need a different route, panels that must travel

Before and after

BeforeAfter
Finding one hour for a panel of threenine emails across several daysslots offered from live calendars, chosen by the candidate
What the interviewer has on the daywhatever was in the invitationCV, role profile, competency questions and scorecard in Teams
When the scorecard is writtendays later, from memoryfrom the card, minutes after the interview
A missing piece of feedbacknoticed when the recruiter chasesan SLA task, then an escalation to the hiring manager

Systems and integrations

We do not add technology to make an architecture look serious. Every element below has a specific job in this process.

Inputs

  • the ATS stage change
  • interviewer calendars in Exchange Online
  • the SharePoint competency question bank

Automation layer

  • UiPath Orchestrator
  • UiPath Robots
  • UiPath Integration Service
  • UiPath Action Center

Target systems

  • the ATS (slots, scorecards, referrals)
  • Exchange Online calendars
  • Microsoft Bookings
  • Power BI

Human touchpoints: the candidate's Bookings page; interviewer brief and scorecard cards in Microsoft Teams; Action Center feedback tasks and hiring-manager escalations

the ATS stage changeUiPath OrchestratorUiPath Robotsthe ATSthe candidate's Bookings page

Technologies used

UiPath Robots + Orchestrator

react to the stage change, run the slot search and the invitation logic, hold the feedback SLA, retries and audit trail

A
UiPath Integration Service (Microsoft Outlook 365 and Microsoft Teams connectors)

calendar and mail actions, cards to interviewers and hiring managers

A
Microsoft Graph (calendars and Microsoft Bookings)

free/busy across the panel and the candidate's self-scheduling page inside your own tenant

A
Microsoft Teams (Adaptive Cards)

the interviewer brief, the scorecard prompt and the hiring manager's pipeline card

A
Microsoft Forms

the scorecard itself, mapped back to the ATS record

A
UiPath Action Center

the SLA task when feedback is missing, escalated to the hiring manager and completed in Teams

A
Workday Recruiting or SAP SuccessFactors Recruiting

requisitions, panels, candidates, stages and the scorecard record

A
Power BI

days per stage, feedback compliance and interviewer load by team

A
Averified product capability (vendor documentation)

Illustrative economic model

What it is worth, with the arithmetic shown.

Illustrative model
364 interviews a month (0.7 of 520) × 24 minutes of coordination≈ 146 h / month
146 h × €32 fully loaded coordinator cost≈ €4,659 / month
× 12 months≈ €55,910 / year
Annual coordinator capacity released (illustrative)≈ €55,910

The calculator holds one of the two clocks in this process. The coordinator's: 520 interviews a month × 24 minutes is 208 hours, of which seven-tenths, 146 hours, can move to the process at €32 a fully loaded hour, which is why the volume below is 364 rather than 520. The interviewer's clock is priced in the next section instead: six minutes of email traffic per interview across 520 interviews is 52 hours a month at €72, about €44,928 a year. Vacancy weeks, agency fees and the candidate who accepted elsewhere sit outside every figure here, and none of this was measured at a client.

Run the numbers on your data

hours released per month
of annual capacity released

An illustrative estimate from your own inputs. It models released capacity; it is not a promise of savings.

Business benefits

  • Time from the interview stage to the offer shrinks, because finding an hour takes minutes of machine work instead of days of email
  • Interviewers arrive prepared, since the brief reaches them in Teams with the CV, the profile and the questions already attached
  • Feedback lands the same day, because the scorecard is one tap from the meeting and a missing one is chased without a recruiter noticing
  • Candidates experience a process that answers, which decides more offers than most hiring teams admit
  • Referrals become a tracked channel with the referrer credited in the ATS, instead of a favour that ends in somebody's inbox

The management view

  • Days per stage, per team and per interviewer replace the impression that recruiting is slow with a number that says where
  • Interviewer load becomes visible, so the same three engineers stop carrying every panel while their colleagues carry none
  • Feedback compliance is measured, which changes behaviour before any policy does
  • Every scheduling and feedback event is logged, so a dispute about who held up a hire is answered from the record

Board-level KPIs

days from interview stage to first interviewdays from interview to feedbackshare of feedback received within one dayinterviews per interviewer per month

Security and governance

An auditor should be able to reconstruct every decision.

  • Candidate data carries a clock. The robot reads only the fields scheduling and briefing need, holds nothing outside the ATS and your Microsoft 365 tenant, and brief cards expire once the interview has happened
  • Calendar access uses Microsoft Graph application permissions limited to the interviewer calendars in scope; robot secrets live in a credential store integrated with Orchestrator, and the robots run in the EU region of UiPath Automation Cloud
  • Scorecards are visible to the panel and the hiring manager only, and are written to the ATS as the record of the decision; Microsoft Purview retention governs the Forms responses
  • Referral cards capture the referred person's consent before any of their data enters the ATS
  • No model reads anything in this process. Every slot offered, reminder sent and escalation raised follows a rule that recruiting can read, change and show to a works council

Why now

01

Engineers and specialists decide within days, and a loop that consumes three weeks of calendar loses candidates to firms that answer in three; speed has become a hiring capability rather than a courtesy

02

The plumbing is no longer a project: calendar availability and Bookings pages sit inside the tenant you already pay for, and modern ATS platforms publish stage changes through an API a robot can subscribe to

03

The money is visible once it is split: €4,659 a month of coordination and €3,744 a month of interviewer time, recurring, before anyone counts a lost candidate

Relevant executive roles

CHRO

Time to hire and candidate experience become measured outcomes instead of anecdotes traded in a leadership meeting

HR Director

Recruiter and coordinator capacity moves from calendars to sourcing and closing, which is the work that actually fills roles

Hiring Manager

Panels arrive prepared, feedback appears the same day, and the pipeline is visible in Teams without a status meeting

Common questions and objections

Our ATS already has scheduling.

Most ATS scheduling stops at the recruiter's own calendar. The robot works across the panel's Outlook calendars, applies your rules to the slots, and creates the Teams meeting, which is the part that consumes the days.

Engineers will ignore reminders the way they ignore everything else.

A one-tap scorecard that appears in Teams minutes after the interview gets filled in far more often than an email two days later, and the SLA escalation handles the remainder without a recruiter chasing.

Candidates will pick inconvenient slots.

The rules decide which slots exist; the candidate only chooses among them. Notice periods, working hours, minimum panel size and buffers are yours to set and change.

When this is not the right solution

  • Hiring volume is a few dozen people a year, where a coordinator with a good template is cheaper than an integration
  • Interviews are mostly single-interviewer phone screens, so there is no panel to reconcile and little coordination to remove
  • The ATS is being replaced within the year, or interviewers do not keep Outlook calendars, in which case availability cannot be read and the flow can only propose, not confirm

A question for the next management meeting

Four hours of interviews spread across three weeks of calendar: who in this company owns the difference, and what has it cost us in accepted offers?

Implementation approach

What we deliver, and what we need from you to start.

We deliver

  • Reconstruction of the last two quarters of requisitions from the ATS, showing where the elapsed days actually went
  • The scheduling rules, the candidate slot flow and the reschedule logic, agreed with recruiters and hiring managers
  • ATS integration for stage events, panel data, scorecard write-back and referrals
  • The interviewer brief, scorecard and pipeline cards in Microsoft Teams, and the feedback SLA in Action Center
  • Referral and open-roles cards on the same plumbing, with the referrer credited in the ATS
  • A pilot with one hiring team end to end, then rollout, training and the funnel report

We need from you

  • API access to the ATS and calendar permissions for the interviewers in scope
  • A pilot hiring team willing to run its requisitions through the new flow
  • The competency question bank, or the decision to build one during discovery
  • The scheduling rules you want enforced: working hours, notice periods, minimum panel size

Stages

Discovery

Requisition timestamps, loop shapes, where days are lost, rules and exceptions

Design

Slot rules, card content, scorecard mapping, SLA and escalation model, security

Build

Robots, ATS integration, Bookings flow, Teams cards and the funnel report

Validation

Dry runs on real requisitions with one hiring team, including reschedules and no-shows

Go-live

One team live with supervision, then team by team

Optimisation

Rules tuned on measured loop times; referral and open-roles cards added

Departmental. Effort follows the ATS API, the number of hiring teams with genuinely different rules, and whether interviewer calendars can be trusted as a source of availability.