Recruiting · Production

AccioJob reduces false-positive assessments by 70% with Retell AI voice invigilator

The problem

AccioJob's AI-driven assessments suffered a false-positive rate as high as 50% due to sophisticated cheating tools, preventing reliable identification of top talent for partner companies. Building an in-house solution to combat cheating was deemed too time- and resource-intensive.

First attempt

Each new assessment AccioJob released was initially effective but candidates rapidly found workarounds, forcing the team to iterate on their platform every few months in a continuous cat-and-mouse game with cheaters.

Workflow diagram · grounded in source
1
Test completion triggers AI call
Trigger
Immediately after the candidate finishes their test, the AI calls them for 5-6 minutes of follow-up questions.
▾ source quote
“Immediately after the candidate finishes their test, the AI calls the candidate to ask follow-up questions for 5-6 minutes”
2
AI targeted follow-up questioning
Ai action
The AI identifies the candidate's highest-scoring question and asks targeted, more difficult follow-up questions.
▾ source quote
“The AI is able to identify the candidates' highest-scoring question and ask targeted, more difficult follow-up questions”
3
Authenticity score generated
Output
AccioJob generates an authenticity score from call data as a key metric to identify whether candidates have cheated.
▾ source quote
“AccioJob generates an authenticity score, a key metric to identify whether candidates have cheated”
4
Talent management system integration
Integration
Retell AI integrates with AccioJob's existing talent management system to evaluate candidates at scale.
▾ source quote
“integrating Retell AI with their existing talent management system, AccioJob is able to evaluate candidates at scale with personalized tests”
5
Self-optimizing false positive reduction
Feedback loop
Early feedback suggests the false positive score is decreasing over time as the system improves.
▾ source quote
“early feedback suggests that the false positive score is decreasing over time”
Reported outcome

Integrating Retell AI reduced false-positive assessment scores by 70%, dropping the false positive rate from 50% to 15%, while delivering 40% client cost savings versus other solutions.
Integration was completed in under two days.

Reported metrics
False-positive assessment reduction70%
Integration time<2 days
Client cost savings vs other solutions40%
False positive rate before integration50%
Show all 8 reported metrics
false-positive assessment reduction70%
integration time<2 days
client cost savings vs other solutions40%
false positive rate before integration50%
false positive rate after integration15%
cost saving vs human interviewer30-40%
partner companies using AccioJob toolsover 20
applications per software engineer role postingmore than 1,000
Reported stack
Retell AItalent management system
◆ Does this fit your context?

Compare to your context

Tell us your scale, team, and constraints. We'll show what changes at your size, what fails at your scale, and whether this case is a fit, needs adaptation, or won't scale to you. Free demo, no signup.

Compare to your context →
~30 seconds · free
Source
https://www.retellai.com/case-study/how-acciojob-saw-a-70-reduction-in-false-positive-assessments-with-retell-ai
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Integrating Retell AI reduced false-positive assessment scores by 70%, dropping the false positive rate from 50% to 15%, while delivering 40% client cost savings versus other solutions.

What tools did this team use?

Retell AI, talent management system.

What results were reported?

False-positive assessment reduction: 70%; Integration time: <2 days; Client cost savings vs other solutions: 40%; False positive rate before integration: 50% (source-reported, not independently verified).

What failed first in this deployment?

Each new assessment AccioJob released was initially effective but candidates rapidly found workarounds, forcing the team to iterate on their platform every few months in a continuous cat-and-mouse game with cheaters.

How is this recruiting AI workflow structured?

Test completion triggers AI call → AI targeted follow-up questioning → Authenticity score generated → Talent management system integration → Self-optimizing false positive reduction.

Related recruiting cases

WHAT TO DO WITH THIS

Now compare it to your context

This case is one data point. Whether its pattern fits you depends on your volumes, your stack, and your exception load — that comparison is the step no case study can do for you.

Browse the Recruiting cases →