How Jooble Cut Time to First Contact with Candidates to 2–3 Minutes with UniTalk

In recruiting, speed wins.

The moment a candidate submits an application is the moment their interest in the job peaks. They may be weighing several offers at once, so even a few hours of waiting can make a difference.

For the Ready2Recruit team, this was an everyday challenge. New candidates kept coming in, and each one needed to be contacted as quickly as possible. But there’s a physical limit to how fast a team can work. People simply can’t handle the entire flow of contacts at once.

That’s where Jooble’s journey toward automating first contact began.

Ready2Recruit, a separate division of the company, helps businesses with initial candidate sourcing. The team receives a request from an employer, finds potential candidates in its own database, reaches out to them, and passes the shortlisted contacts on to the client.

Ready2Recruit focuses mainly on roles where employers actively compete for people who are ready to start work.

So the question wasn’t just how many candidates the team could process. What really mattered was how quickly a candidate would get their first call after appearing in the database.

At first glance, the solution seems obvious: more candidates means more people to make calls.

But for Ready2Recruit, that wasn’t a one-size-fits-all answer. The number of employer requests fluctuates, and the team’s workload can spike due to vacations, sick leave, or other circumstances.

On top of that, the task itself was fairly mechanical. Contact a large number of candidates quickly, ask a few basic questions, figure out whether the person is a fit, and pass the result along.

So the team was looking for a way to speed up first contact without constantly expanding headcount.

Before choosing a solution, Jooble considered around 15 contact center options.

In the end, they chose UniTalk.

The first step was the Predictive Dialer.

It automatically calls through the candidate database and connects an operator only with those who pick up.

It may seem like a small change. But it removes manual dialing and waiting for a connection from the recruiter’s day. A person joins the call only when there’s already a candidate on the line.

This is how the team automated the first and most repetitive part of the process.

Next came a voice bot.

It handles the first conversation with the candidate and helps filter out those who are no longer looking for work or don’t meet the basic criteria.

The recruiter is left with the candidates worth pursuing.

The next stage was a different level of automation altogether.

The Voice AI Agent doesn’t just call a candidate and run through a fixed list of questions. It holds a conversation, responds to answers, asks screening questions, and logs the outcome of the call.

In other words, it’s no longer just the call that’s automated, but the entire initial communication process.

This is where it got interesting.

Connecting an AI agent isn’t enough for it to start talking to candidates flawlessly right away. Together with UniTalk, the Jooble team tested scenarios, refined scripts, analyzed conversations, and adjusted the agent’s logic.

Essentially, the AI went through the same path as a new hire. It needed the process explained, its behavior checked in different situations, and its work gradually polished.

That’s why, for Jooble, working with UniTalk went beyond simply plugging in a service.

Making the conversation sound natural was a separate challenge.

The AI agent was given a name: Daryna. The team worked not only on what she says, but on how she says it.

They fine-tuned stress, pauses, pace, intonation, even breathing. They also added light background noise so the voice wouldn’t sound too sterile.

Sometimes the effect of these adjustments shows best not in numbers, but in the reaction of the person on the other end of the line:

Once initial communication no longer depended solely on the team’s manual work, automation started delivering results across the entire process.

Every month, Ready2Recruit makes 150,000–250,000 dials and works with 12,000–32,000 unique candidates.

From this flow, 120–360 candidates are selected each month for further work with clients.

At that volume, even a few minutes spent on one repetitive action add up to a significant share of the team’s time.

That’s why it was important for Jooble to automate exactly those stages where human involvement wasn’t essential.

One example shows clearly why speed matters here.

One of Jooble’s clients needed to fill vacancies quickly in several towns near Kyiv.

The Voice AI Agent called candidates in these towns and, within two days, helped find 20 people who met the job requirements.

In recruiting, getting through isn’t everything. After the conversation, you need to understand what actually happened: Is the candidate interested? Do they meet the job requirements? Should they be passed on to a recruiter?

So the next question was: how to avoid bringing manual work back at the call analysis stage?

Alongside the Voice AI Agent, Jooble uses Speech Analytics. It automatically analyzes conversations and classifies candidates.

According to the case data, around 80% of suitable candidates are classified automatically, without anyone manually listening to calls.

This way, automation covers not only the call itself, but also everything that happens with its outcome.

Ready2Recruit needed automation primarily for the initial stage of working with candidates.

The Voice AI Agent handles calls, initial screening, and logging the outcome of each conversation. This allows the team to work with a large flow of candidates without the same volume of manual work.

In practice, automation reduced human involvement exactly where the work was most repetitive: initial outreach calls and candidate screening.

This made it possible to rebuild Ready2Recruit’s workflow and significantly speed up first contact with candidates.

The main change for Jooble is the speed of first contact.

Instead of waiting until the next business day, a candidate can now get their first call within 2–3 minutes.

Speed, however, is only part of the result. Automation now covers the entire initial process: from dialing and the first conversation to candidate screening and analysis of the results.

And the difference in speed matters in practice. According to Jooble, the likelihood of a candidate answering varies by more than 20% depending on how quickly the first call is made.

For a process where reaching a candidate at the peak of their interest is critical, this is more than just a convenience.

Jooble’s story isn’t about AI simply calling candidates instead of people.

The team automated the process step by step: first dialing, then initial voice communication, and finally a full conversation with the candidate and analysis of the results.

Along the way, the AI itself had to be adapted to the real process: testing scenarios, analyzing conversations, and refining its logic.

As a result, time to first contact dropped to 2–3 minutes, and the team gained a tool that helps handle a large flow of candidates without a proportional increase in manual work.

UniTalk’s Voice AI Agent can automatically call candidates, run initial screening, and log the outcome of each conversation.

Book a demo and see how this scenario could work in your process.