Perry Ellis · Yotpo
1.4% — conversion rate
Independently reportedRetail & e-commerce · Customer support · NLP · Scaled
Customer support
VRK (Central Electoral Commission of the Republic of Lithuania) · Tilde
Vendor-reportedPublic sectorLithuaniaIn productionPublic sectorNLPLive 2024
Vendor-reported. The customer is named and the numbers are quoted from the source page, but the account comes from the vendor. No independent confirmation.
Reported by tilde.ai (vendor self-report). Checked against the source page on 2026-09-06. 6 of 6 figures below appear on that page word for word.
The problem
During election time, it was not physically possible for VRK staff to deal with organizational issues and communicate with voters/campaigners simultaneously, leading to staff overload.
What was deployed
- Vendor
- Tilde
- Products
- VRK chatbot
- Technique
- NLP
- Build or buy
- Built in-house or to order
- Deployment
- SaaS
- Data used
- large knowledge base; wealth of information relevant to the public
- Scale
- 24/7 availability
What changed
Each row is quoted from the source. Figures we could not find on the page in those words are marked — they are kept, not deleted, so you can judge them.
almost 5000visitorssince launch on 18 March
“The VRK chatbot, which was launched on 18 March, has had almost 5000 visitors, 4422 conversations and 9852 messages during this period.”
Quoted word for word from the source
4422conversationssince launch on 18 March
“The VRK chatbot, which was launched on 18 March, has had almost 5000 visitors, 4422 conversations and 9852 messages during this period.”
Quoted word for word from the source
9852messagessince launch on 18 March
“The VRK chatbot, which was launched on 18 March, has had almost 5000 visitors, 4422 conversations and 9852 messages during this period.”
Quoted word for word from the source
about 103messages per dayfrom 18 March until early voting in presidential elections
“from 18 March until the early voting in the presidential elections, the chatbot received on average about 103 messages per day”
Quoted word for word from the source
500messages per dayafter early voting started
“after the early voting started, this number began growing rapidly, reaching 500 messages per day”
Quoted word for word from the source
1,738messages per dayduring the election week
“during the election week, the VRK robot handled many more queries, answering 1,738 messages per day on average.”
Quoted word for word from the source
ROI as stated: the robot has already paid for itself.
“During election time, together with other colleagues from VRK, we start performing administrative work from around 6 p.m. when the calls from voters stop. It is simply not physically possible to deal with the organisational issues of the elections and communicate with voters or political campaigners at the same time, despite the fact that both functions are very important. This is why the idea was born to deploy an intelligent AI chatbot, as it would allow voters to get the answers they need as quickly as possible.”
Difficulties and limits
- ensuring accuracy of information to avoid misleading answers, especially for sensitive election-related data
- cybersecurity concerns to protect the chatbot from aggressors
- chatbot can only handle queries in Lithuanian right now
Sources
- tilde.ai — vendor self-report, 2024-07-09
Similar deployments
Innovative Dining Group · Revmo AI
1200+ — Reservations booked by AI
Vendor-reportedTravel & hospitality · Customer support · NLP · In production · 2025
Houston Methodist · ActiumHealth
18000000 — staffing capacity created
Vendor-reportedHealthcare · Customer support · LLM · In production
Luxury Bath NJPA · Hatch AI
52% — Overall set rate
Vendor-reportedConstruction & real estate · Customer support · AI agents · Scaled
Radiant Credit Union · interface.ai
60% — Asset growth absorbed without contact center team expansion
Vendor-reportedBanking & finance · Customer support · LLM · Scaled · 2022
Talkmobile · LivePerson
92% — Customer Satisfaction Score (CSAT)
Vendor-reportedTelecom · Customer support · NLP · Scaled · 2014