DVET Maharashtra - Case Study
How DVET Maharashtra automated 50,000+ student queries with a bilingual AI chatbot
DVET Maharashtra deployed BotPenguin’s bilingual chatbot in Marathi and English across 1.9M+ annual website visitors. Student satisfaction rose by 52% and ITI admissions conversion improved by 35%.
Figures are reported by DVET Maharashtra and reflect operational data measured before and after deployment. Individual results vary.
Industry:
Government Education
Visitors Served:
1.9M+ annually
Geography:
Mumbai, Maharashtra, India
Established:
1948
Key feature used:
Bilingual website chatbot (Marathi & English)
Website:
dvet.gov.inThe Challenge
Manual student query handling at scale was failing during admission season
DVET Maharashtra oversees ITI admissions and vocational training across six regional offices. With 1.9M+ annual visitors and a sharp query surge every June to August, manual handling created delays and missed engagement, and Marathi-speaking students hit a language barrier that English-only support could not solve.
Admission-season surge broke manual capacity
Thousands of students queried the website simultaneously during June to August about admission dates, eligibility, and documents, faster than staff could keep pace.
Repetitive queries consumed staff time
Most questions had standard answers already on the site, so complex cases that needed human attention were delayed.
Bilingual accessibility gap excluded students
Rural and semi-urban Marathi speakers were underserved by English-only support, a barrier a state-level directorate could not accept.
No after-hours support during admissions
Students researching ITI admissions outside office hours got no response, and dropped off before completing applications.
Inconsistent handling across offices
Six regional offices and a central directorate handled queries differently, so students received inconsistent information depending on where they asked.
Operational stakes
- Admission-season queries went unanswered
- staff capacity was consumed by automatable questions
- and there was no 24/7 guidance during the high-stakes ITI admission window
Why BotPenguin
DVET needed a platform that could serve a state, not just a website
Most chatbot platforms are designed for businesses. DVET needed one that served a multilingual state-level audience, held up under peak-season traffic, and delivered accurate guidance when student decisions depended on it.
The platform adapted to the audience
Language detection and native Marathi response were built into the core logic, so accessibility was not an add-on.
Accuracy came from training
The chatbot was trained on DVET’s actual ITI admission criteria and scheme eligibility rules, not a generic education template.
The infrastructure held when it mattered
Peak admission season was handled without service degradation during the period of highest consequence.
Staff were redirected, not replaced
The bot resolved what could be automated and handed off what could not, moving the team from managing volume to managing complexity.
| Feature | How it was used |
|---|