evidoria

← Back to browse

Good practice Imported

Iti — Statistics Estonia's AI Virtual Assistant for Public Data Enquiries

Estonia · Tallinn · See the Estonia profile · See the Tallinn profile

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

Statistics Estonia's chatbot Iti, built with local start-up AlphaBlues and launched in April 2019, answers citizens' and data-providers' questions about population, wages and prices in natural language, freeing staff to handle complex enquiries.

~5,000
Information requests handled annually by 2019 (by 2019)
~1,000
Information requests handled annually two decades earlier
~1 year
Development period before Iti's launch
Iti — Statistics Estonia's AI Virtual Assistant for Public Data Enquiries

Details

Maturity
Established
Promoter
Statistics Estonia (Statistikaamet)
Period
2018-present
Region (NUTS)
EE00
Keywords
public statistics, citizen services, data access

Context

Statistics Estonia (Statistikaamet) introduced a live-chat feature on its website in 2017, and the volume of enquiries handled through it grew rapidly, reaching roughly 5,000 information requests a year by 2019, up from about 1,000 a year two decades earlier, per the agency's own account.

Objectives

To manage this growing load without expanding front-line staff, Statistics Estonia partnered with Tallinn-based start-up AlphaBlues to build 'Iti', an NLP chatbot intended to answer common questions about population figures, the gender pay gap, average wages and consumer-price trends, and to help data providers navigate the eSTAT electronic reporting portal.

Activities

Iti was trained on real user questions over roughly one year before its launch in April 2019, and it escalates anything it cannot answer to a human specialist.

Conclusions

The EU Joint Research Centre's AI Watch catalogue and Statistics Estonia's own press material describe Iti as one of the first AI deployments in an EU public statistics office, but neither source publishes a post-launch accuracy rate, containment rate or user-satisfaction score, so the evidence here is descriptive rather than measured, which should temper claims of impact.

Implementation
Indicative cost
Low (< €50k)
Time to results
Medium (1–3 years) — Live chat introduced 2017; Iti developed over roughly one year and launched April 2019.
Staffing & skills
Statistics Estonia (Statistikaamet), AlphaBlues (Tallinn-based start-up, chatbot developer)

Conditions for success

  • NLP chatbot trained on roughly a year of real user questions before launch
  • Escalation path to a human specialist for unanswered queries
  • Integration with the eSTAT electronic reporting portal for data providers

Common failure modes

  • No post-launch accuracy, containment or user-satisfaction metrics have been published

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.

Attachments

Similar practices you may find useful