On-call interpreters for police, hospitals and labour offices serving migrants
TL;DR Parleyline gives a police station, government hospital or labour office one number to call when the person in front of them speaks another language. Machine translation handles routine exchanges, and a human interpreter joins for statements, consent and anything with legal or medical weight. State departments pay a yearly contract.
India's internal migrants often end up where nobody at the counter speaks their language. Kerala alone had more than 25 lakh migrant labourers by 2015, and police there were taking informal Hindi lessons because they couldn't question or help them[1]. Researchers studying Kerala's migrant workers call language a large and overlooked barrier to healthcare, one that gets in the way of diagnosis and of sticking to treatment[2].
Both sides pay for it. A worker can't explain an injury or give a clear statement, and an officer or doctor makes decisions on half-understood answers. A 2013 paper in the National Medical Journal of India noted that the usual fixes, like on-site or telephone interpreters, had barely been tried in India[3].
What it does, in regulatory terms: language interpretation only. Parleyline does not give legal or medical advice and does not record statements on anyone's behalf. Officers and doctors stay responsible for what they write down.
The service is a phone number and a simple app. A desk officer picks the language pair and the purpose. For routine stuff (directions, document lists, appointment times) Parleyline uses machine speech translation built on public Indian-language models such as Bhashini[4]. For a police statement, a hospital consent form, a discharge explanation or a labour dispute, a trained human interpreter joins the call within minutes, on a three-way line or video.
Each call ends with a short log of the language, the duration and whether a human joined. No content is stored unless the department asks and the person agrees. The interpreter pool comes from bilingual graduates in the source states, trained on police, hospital and labour vocabulary.
The user is the officer or clinician, and the person being helped pays nothing. The payer is a state department - police, health or labour, usually in states that receive lots of migrants. Large employers and construction firms with welfare obligations are a second buyer. There's no public Indian price for on-call interpretation, and that's the main thing to test.
Machine speech translation for Indian languages is now a public service. Bhashini, launched by MeitY in July 2022, offers more than 300 models covering speech recognition, translation and speech synthesis[4][5]. That lets a small human pool focus on the calls that really matter. Departments also have growing digital budgets and a record of language problems that Hindi lessons won't fix[1].
There is no published figure for interpretation spending by Indian public services. The narrow signal is demand - over 25 lakh migrant workers in one destination state by 2015[1], dealing with police, hospitals and labour offices that mostly don't share their languages[2]. On the adjacent side, one large US interpreting firm runs on-demand interpreting across more than 240 languages for healthcare and public safety clients[6].
Bhashini supplies the translation models but not a staffed service for departments[4]. Large interpreting firms show the model works abroad, for hospitals and public safety[6]. In India, police lean on staff who happen to speak the language, or on the worker's employer or a co-worker, and the conflicts there are obvious. The gap is a neutral, trained, on-call interpreter line paid for by the state.
| Difficulty | Moderate (45/100) |
| Starting capital | ₹80 L–1.4 Cr |
| Time to launch | 6–9 months |
The hard parts are the human pool and procurement. You need reliable interpreters in Odia, Bengali, Assamese and other source languages on call at night, and they have to stay neutral. Machine translation of accented speech in a noisy police station will get things wrong, and you have to route around that. Government contracts also take ages to close.
No licence to provide interpretation, but police and hospital work carries real legal weight.
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
Don't trust machine translation for police statements, medical consent or anything a court may read later. Parleyline interprets only and gives no legal or medical advice, and accented speech in a noisy room will trip a model up.
Want to work out the AI layer for Parleyline? Talk to Yash →
I think this is a worthwhile business for a founder with a background in public health, policing or language services who can recruit and supervise interpreters. If you're hoping machine translation alone will do the job then it's a poor fit. Prove first that one district's police and hospital use the line 10 times a day.
Plan for roughly ₹80 L–1.4 Cr, most of it on building and paying an on-call interpreter pool, plus the long wait for state contracts.
No licence to interpret. You do need DPDP compliance, legal advice before police work and a written code for interpreters in hospitals.
Yes, it's an established model abroad, with large interpreting firms serving public bodies[6]. Check local rules on interpreter qualifications and how interpreted statements are used in court.
The free Parleyline validation kit has the line-by-line capital breakdown, the effort scoring, the data-protection and procurement checklist, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for officers and clinicians, and 5 assumptions to test. Download the kit →
Thinking about building Parleyline or something close to it? The make-or-break bit is a reliable night-time interpreter pool and a state department willing to pay for it, and I can help you work out how to land both. 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.