evidoria

← Back to browse

Good practice Imported

Tengai the Interview Robot — Upplands-Bro's Bid to De-Bias Municipal Recruitment

Sweden · Upplands-Bro · See the Sweden profile

Evidence: Descriptive / self-reported Top 75% 47/100 · Ask Evidence Copilot about this practice

Upplands-Bro became the first municipality worldwide to screen a hire with the Tengai AI interview robot, stripping gender, age and appearance cues from first-round interviews; a 2019 pilot of ~200 candidates preceded its first robot-screened hire.

~200
Trial candidates screened by Tengai before the first live hire
76 %
Trial candidates who said they would be interviewed by the robot again

Details

Maturity
Pilot
Promoter
Upplands-Bro Municipality (Upplands-Bro kommun), with TNG Group and Furhat Robotics
Period
2019
Region (NUTS)
SE110
Keywords
human resources, recruitment, bias mitigation, social robotics

Context

Sweden's Upplands-Bro municipality partnered with recruitment firm TNG and the robot's maker, Furhat Robotics, to trial Tengai Unbiased, a physical AI interview robot, aiming to remove unconscious bias — gender, age, appearance, accent — from first-round candidate screening.

Objectives

Tengai asked every candidate the same fixed, competency-based questions with no evaluative body language or follow-up probing, so that transcripts could be scored on skills content alone before human recruiters carried out reference checks and in-person interviews with the shortlisted candidates.

Activities

Developer testing across roughly 200 candidates found most adapted quickly to the robot interviewer, and in July 2019 Upplands-Bro used Tengai for its first live public-sector hire — a digitalisation coordinator — with deputy HR chief Havva Ilhan calling the process smooth.

Results

TNG separately reported that 76% of trial candidates said they would be interviewed by the robot again.

Conclusions

The municipality itself cautioned that the robot 'probably cannot completely replace human recruiters' and used it only as a first-round filter. No independent, peer-reviewed study of whether Tengai actually reduced biased hiring outcomes at Upplands-Bro has been published — the adoption and satisfaction figures reported come from the vendor and the municipality, not an outside evaluator.

Implementation
Indicative cost
Medium (€50k–€500k)
Time to results
Short (< 1 year) — Developer testing across roughly 200 candidates in 2019 preceded the first live public-sector hire in July 2019.
Staffing & skills
Upplands-Bro Municipality HR team (incl. deputy HR chief), TNG Group (recruitment firm partner), Furhat Robotics (robot developer)

Conditions for success

  • Fixed, competency-based question set applied identically to every candidate
  • No evaluative body language or follow-up probing from the robot
  • Human recruiters retained for reference checks and in-person interviews of shortlisted candidates

Common failure modes

  • No independent, peer-reviewed study confirmed the robot reduced biased hiring outcomes
  • Adoption/satisfaction figures reported are vendor (TNG) and municipality self-reports, not from an outside evaluator
  • Municipality itself acknowledged the robot 'probably cannot completely replace human recruiters'

Where it fits

Governance type
municipal government
Scale
single municipality / single hire

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful