Phone role-play tests in Indian languages for frontline job candidates
TL;DR Roleplaycall tests whether a job candidate can calm an annoyed customer, in the language they'll actually work in. The candidate takes an eight-minute phone call with a simulated customer and the call is scored against a job rubric. Skilling partners and frontline employers pay per candidate.
India trains and certifies people at scale, yet a lot of them still don't land or keep jobs. By mid-June 2023, PMKVY schemes had assessed 12.45 million candidates[1]. In 2022, 65.7% of India's unemployed were young people with secondary education or more, up from 54.2% in 2000[1]. The ILO also found that youth with formal vocational training had higher unemployment than those trained on the job, and it named skill mismatch as one reason[1].
For counter, delivery, call-centre and front-desk roles, the gap is often how a person handles a difficult customer. Most screening tests for this are written or in English, while the job is done in Hindi, Marathi or Tamil. That mismatch is silly.
What it does, in regulatory terms: a skills assessment that informs hiring and training decisions, and does not make them. Calls are personal data under the DPDP Act and Rules, so candidates give clear consent first[2]. Roleplaycall does not infer personality, emotion or honesty from a voice.
A training partner or recruiter picks a role, such as quick-commerce rider, retail floor or hospital front desk, and a language. The candidate dials in or answers a call, and a software customer complains about a late order or a wrong bill and pushes back once or twice. Roleplaycall scores the recording on a short rubric - did they understand the issue, stay polite, offer a real fix and close the call? Trained human raters score a sample to keep the scores honest. Trainers get a two-minute clip with marks and a coaching note.
I reframed the seed's immersive assessments. Candidates don't have VR headsets but they do have phones. Testing one job task in the working language is also far easier to defend than scoring "soft skills" in general.
The candidate takes the test free and two buyers pay. Placement-linked training partners want proof before they send a batch to interviews. High-volume employers and staffing firms want to screen out mismatches before onboarding costs pile up. Big assessment vendors quote enterprise prices privately[3][4], so my test price, a flat fee per assessment well below one day's onboarding cost, is an assumption to check.
Speech scoring in Indian languages now runs at government scale. One AI non-profit reports more than 15 million oral reading assessments across two states, as of January 2026[5]. Phone-based testing is familiar too, with one mobile English test from a UK cultural body reporting over 15 million tests taken[6]. And the DPDP Rules notified in November 2025 make consent duties for voice data clear[2].
I found no public figure for spending on soft-skills or communication assessment in India. The narrowest volume anchor is the 12.45 million PMKVY candidates assessed by June 2023[1], though those were certification tests and not hiring screens.
Adjacent - one hiring startup that recruits blue-collar workers through chatbots, claims about 20,000 placements a month[7], which shows how many frontline hires already move through tech-led funnels.
Big online assessment vendors sell communication skills and situational judgement tests to employers[3]. Global talent-assessment firms offer call-centre simulations and language evaluation for volume hiring[4]. A free mobile test checks English on a phone[6]. Chatbot hiring platforms handle sourcing rather than skill proof[7].
The gap is a cheap, per-candidate role-play in the candidate's working language, built for training partners as much as for employers.
| Difficulty | Moderate (39/100) |
| Starting capital | ₹50–90 L |
| Time to launch | 5–7 months |
A believable customer voice is the easy half. The hard half is proving the score means something, that it agrees with experienced supervisors and predicts who lasts three months in the job, and that needs a validation study with real employers. Speech recognition on cheap phones, with noise and code-mixed Hindi-English, will also drag accuracy down.
Selling a hiring or practice screen needs no licence. Voice recordings bring data duties, though.
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
Don't let AI reject candidates on its own, and never use it to judge personality, emotion or honesty from a voice. Accents and poor phone lines will skew scores.
Want to work out the AI layer for Roleplaycall? Talk to Yash →
I'd back a founder from frontline hiring, BPO operations or a placement-linked training partner who can get employers into a validation study. If you want to sell an AI personality test then this isn't for you. Prove one thing first - Roleplaycall scores line up with supervisor ratings for 200 candidates in one role and one language.
Plan for roughly ₹50–90 L, mostly on voice technology in several Indian languages, rubric design and employer sales.
No licence for a hiring or practice screen. You need DPDP-compliant consent for recordings, and NCVET recognition only for certified NSQF assessments.
Yes, and the frontline hiring problem is common. Check local data protection law and any rules on automated hiring decisions before you launch.
The free Roleplaycall validation kit has the line-by-line capital breakdown, the effort scoring, the consent and fairness checklist, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for training partners and hiring managers, and 5 assumptions to test. Download the kit →
Thinking about building Roleplaycall or something close to it? The make-or-break bit is showing employers that the score predicts how people actually do on the job, and designing that study well is something I help founders with. If you need a hand, you know where to find me. Talk to Yash →
For illustration only. Names are fictional and ideas hypothetical; sourced facts are real but may be out of date. Nothing here is legal, financial or professional advice, and Yash Khatri is not personally liable for any use of it. Read the full disclaimer.