Skip to content
DT Nova
Services

Contact center scheduling software that plans to the interval

Contact center scheduling software has one job that a general rota tool cannot do: put the right number of agents with the right skills on every 15 or 30 minute interval, so customers are answered within your service level without paying for idle time. We build it around your platform, your queues and your working time rules: forecasts per interval, staffing by Erlang C or simulation, shrinkage, shifts and breaks, intraday changes and an app for agents. Your company owns the code.

  • You see the result before it goes live
  • Fixed price for the agreed scope, in writing before work starts
  • Source code, domain and accounts in your company's name

Request a quote

Why contact center scheduling is different

A shop or a clinic plans by the shift. A contact center plans by the interval, because demand moves by the quarter hour: a Monday morning peak, a lull after lunch, a spike when invoices go out or a service goes down. Being two agents short for one hour can push waiting times from seconds to minutes, while being two agents over all day is a cost nobody sees on the dashboard. On top of that come skills and languages, several channels with different rules, agents who handle several chats at once, breaks and training that must be placed where they hurt least, and working time rules that apply to every shift.

That is why a contact center needs scheduling software that starts from the forecast, not from a blank rota. If you plan shifts for other kinds of teams, see employee scheduling software; for the full loop of forecasting, time and attendance, absence and labour cost across the whole business, see workforce management software.

From forecast to schedule

  1. Forecast. Contact volume and average handle time per queue, channel and interval, built from your platform's history and adjusted for seasons, holidays, campaigns and known events.
  2. Required staff. For each interval, the number of agents needed to meet the service level target, such as 80 percent of contacts answered within 20 seconds, calculated per queue.
  3. Shrinkage. Breaks, training, meetings, coaching, absence and leave are added on top, because an agent on the schedule is not always an agent available for contacts.
  4. Shifts and breaks. The system builds shift patterns that cover the requirement, respects contracts and rules, and places breaks and lunches where the forecast has room.
  5. Publishing. Agents see the schedule in the app weeks ahead, request changes and swaps, and supervisors approve.
  6. Intraday. During the day, actual volume is compared with the forecast, and the plan is adjusted: moving training, offering extra hours, or reassigning skills.

A worked staffing example

Erlang C is the classic staffing formula for queues where customers wait for the next free agent. It takes the forecast volume and handle time for an interval and gives the probability that a customer has to wait, and from that the service level for a given number of agents. The numbers below are an illustration, not a benchmark.

Input or resultValue
Contacts forecast in a 30-minute interval120
Average handle time300 seconds
Workload20 agent-hours per hour (20 Erlangs)
Service level with 23 agents66 percent answered within 20 seconds
Service level with 24 agents77 percent
Service level with 25 agents85 percent, which meets an 80/20 target
Agents to schedule if shrinkage is 30 percent36 (25 divided by 0.7, rounded up)

Two things follow. First, the step from 23 to 25 agents changes the service level from poor to good, which is why interval accuracy matters more than daily totals. Second, the formula has limits: it assumes nobody gives up while waiting, so it slightly overstates the need, and it does not describe chat agents with several conversations at once. For those queues we use simulation or adjusted models, and the schedule always shows which method each queue uses, so your planners can check it.

Intraday management and adherence

A schedule made two weeks ago is wrong by ten in the morning on some days. Intraday management compares actual contacts and handle times with the forecast every interval and shows where the gap is heading. Supervisors see the options: delay a training session, release agents early on a quiet afternoon, offer voluntary overtime to agents who opted in, or move agents with the right skills between queues. Adherence, meaning whether agents are in the state the schedule expects, is shown to supervisors as a team view with the reasons that agents record, not as a ranking that invites gaming.

Platforms such as Amazon Connect show what is possible here: AWS documents short-term forecasts for scheduling up to 18 weeks ahead, long-term forecasts for capacity planning up to 64 weeks ahead, and intraday forecasts updated every 15 minutes for queues with enough history, at least 2,000 unique contacts per week per queue and channel for the last five weeks. (Read on 11 November 2026.) When your platform offers such data, we use it; when it does not, we build the forecast from the raw history.

Rules worth writing down first

Every shift the system proposes is checked against the rules your HR team approves. In the EU, the Working Time Directive sets a minimum daily rest of 11 consecutive hours in every 24 and an average working week of no more than 48 hours including overtime; national law and collective agreements add breaks, evening and weekend supplements and limits on split shifts. Contact centers that run across countries or through outsourcing partners need those rules per site and per contract.

A second rule concerns analytics. The EU AI Act, Article 5, prohibits AI systems that infer the emotions of a natural person in the workplace, except for medical or safety reasons, and the prohibitions have applied since 2 February 2025. (Both sources read on 11 November 2026.) Scheduling has no need for emotion analysis of agents, and ours uses none. We do not give legal advice; your advisers confirm the rules and we turn them into checks the system runs every time.

The app for agents and supervisors

Agents see their schedule weeks ahead, request leave and preferred shifts, swap with qualified colleagues, bid for open shifts and get a notification when something changes. Supervisors approve from the same app and plan on a laptop. Fairness rules, such as rotating early and late shifts or weekends, are part of the engine, so they do not depend on who shouts loudest. The app works on iPhone, Android and in the browser.

Scheduling is half of the capacity question; the other half is how much work each contact takes. Sorting, routing and drafting replies before an agent opens a ticket is covered by our customer service automation page.

Built-in WFM module, suite or custom?

SituationPlatform's own module or a WFM suiteCustom scheduling
One platform, standard queuesUsually the right choiceRarely worth building
Several platforms or outsourcing partnersEach with its own planOne forecast and one plan across all of them
Collective agreement rules and supplementsBasic settingsYour rules as written and approved
Back-office work planned with the same agentsOften outside the toolContacts and back-office tasks in one plan
Payroll export in your formatThe vendor's integration listBuilt for your payroll system
Licence cost per agentGrows with every seatRunning cost follows use, not seats

Suites such as NICE, Verint and Calabrio, and the modules inside Genesys Cloud or Amazon Connect, are mature products. If one of them fits your rules, keep it, and we will say so in the first reply to your brief.

How we build it

  1. Planning workshop. Your planners show how forecasts and schedules are made today, including the spreadsheets next to the tools.
  2. Scope and fixed price in writing. Queues, channels, rules, integrations and what the first phase leaves out.
  3. Forecast on your history. Compared with what actually happened in past weeks, so accuracy is measured before anything is scheduled.
  4. Parallel run. For a few schedule periods, the new plan runs next to the old one, and rules are tuned.
  5. Go live and measure. Service level, occupancy, overtime and planner hours, against the numbers recorded at the start.

You own it

As a standard term, the source code is in a repository in your company's own account, and the database and hosting are in your company's name in an EU region. Agent data is processed under a written data processing agreement. You can keep us for maintenance, run it yourself or hand it to another team. For dashboards on top of the same data, see dashboard development; for the wider system, see custom software development. See prices.

Questions we get

Does it connect to our contact center platform?

Yes, through the platform's own interfaces. Genesys Cloud, Amazon Connect, Five9, Talkdesk and most other platforms expose historical volumes, handle times and agent states. We read those for the forecast and for adherence, so nobody exports spreadsheets by hand.

Is Erlang C still good enough for staffing?

For voice queues with patient customers, it is a sound starting point and easy to explain. It assumes nobody hangs up while waiting, so it tends to overstaff slightly, and it does not handle chat agents working several conversations at once. We use it where it fits and simulation or adjusted models where it does not, and we show which one each queue uses.

Can agents swap shifts and choose breaks?

Yes, within the rules. Agents trade shifts, bid for open shifts and request leave in the app; the system checks skills, working time rules and coverage before a supervisor approves. Breaks can be placed automatically where the forecast has room.

We already have a WFM module in our platform. Why build?

Often you should not. The built-in modules are good for the common case. Custom scheduling pays off when you have several platforms, outsourcing partners, collective agreement rules the module cannot hold, or planning that also covers back-office work and field staff.

Does it monitor agent emotions or sentiment?

No. The EU AI Act has prohibited AI systems that infer the emotions of people at the workplace since 2 February 2025, except for medical or safety reasons. Our scheduling uses volumes, handle times, skills and agent states, not emotion analysis of agents.

Who owns the system and the data?

Your company. The code is in your own repository, the hosting and database are in your company's name in an EU region, and agent data is processed under a written data processing agreement.

Get a written quote

Tell us in a few lines what you need. We reply within one working day with a written quote: scope, timeline and a fixed price or an hour estimate. Asking is free and commits you to nothing.

See every price · Contact details

Request a quote