Government & Public SectorConversational AIPublic Cloud

German Federal Employment Agency deploys cloud-based AI with Privatemode

German Federal Employment Agency (Bundesagentur für Arbeit)· GermanyPrivatemode · Llama 3.3 70B

The German Federal Employment Agency (Bundesagentur für Arbeit, BA), with over 110,000 employees and one of the largest public authorities in Germany, needed an internal knowledge management AI chatbot but a local AI deployment would have been very time-consuming. It tested Privatemode, the confidential AI service from Edgeless Systems, which encrypts data during storage, transmission and processing. Using the Privatemode API, the BA set up a chatbot that answers questions from internal databases of its Cloud Center of Excellence, accessing the open-source Llama 3.3 70B model and scaling without owning GPU infrastructure, while keeping user data such as AI prompts protected against access by cloud, model, or service providers.

Overview

The German Federal Employment Agency (Bundesagentur für Arbeit, BA), with over 110,000 employees and one of the largest public authorities in Germany, needed an internal knowledge management AI chatbot but a local AI deployment would have been very time-consuming. It tested Privatemode, the confidential AI service from Edgeless Systems, which encrypts data during storage, transmission and processing. Using the Privatemode API, the BA set up a chatbot that answers questions from internal databases of its Cloud Center of Excellence, accessing the open-source Llama 3.3 70B model and scaling without owning GPU infrastructure, while keeping user data such as AI prompts protected against access by cloud, model, or service providers.

The challenge

The BA operates one of the largest IT landscapes in Germany, with over 100 IT systems and around 15,000 applications for unemployment or child benefits submitted daily. At the same time, the BA faces a significant shortage of skilled workers, as around 38% of its workforce will retire within the next decade. An AI service for internal knowledge management needed to be implemented and tested quickly, but a local AI deployment would have been very time-consuming, so a secure cloud-based LLM solution was to be used, while ensuring sensitive personally identifiable information is processed securely and in compliance with data protection regulations.

The solution

As part of its AI and multi-cloud strategy, the BA evaluated the use of confidential computing and tested Privatemode, the confidential AI service from Edgeless Systems, for its desired knowledge management solution. Privatemode ensures data is encrypted not only during storage and transmission but also during processing. Using the Privatemode API, the BA set up a chatbot capable of answering questions using information from internal databases of its Cloud Center of Excellence, accessing the open-source Llama 3.3 70B model and scaling without needing its own GPU infrastructure, while keeping user data such as AI prompts protected against access by cloud, model, or service providers.

Conversational AILarge Language Models

Reported business value

With Privatemode AI, the BA very quickly set up a secure, scalable AI chatbot capable of answering questions using information from internal databases of its Cloud Center of Excellence, scaling without the need for its own GPU infrastructure while ensuring maximum data protection.

Sources

Open any source and check the claim yourself — that is the point of the register.

This record was researched and written with AI assistance, and its claims were checked against the sources above. (EU AI Act art. 50 transparency notice.)

Related entries

Other government & public sector entries in the register.

All entries
Government & Public SectorIntelligent AutomationPublic Cloud

GovTech unlocks data insights to improve nationwide services with Databricks

Singapore's Government Technology Agency (GovTech) migrated from an on-premises dashboarding system to the Databricks Data + AI Platform on AWS with Unity Catalog and Delta Lake, cutting dashboard creation time from 90 to 30 days, democratizing data across 50% of corporate divisions in the first year, and saving 8,000 labor hours annually.

96/100HighPrimary source
GovTech· SingaporeDelta Lake · Lakeflow Connect · Lakeflow Jobs +1
Government & Public SectorConversational AIPublic Cloud

Estonia rolls out Bürokratt, an AI-guided virtual assistant network for public services

Bürokratt is a network of chatbots deployed on Estonian public sector institutions' websites, letting people obtain information from institutions and use public and information services via virtual assistants. It is a state-created, AI-based digital assistant that helps institutions deliver modern, efficient, around-the-clock customer service using large language models.

96/100HighPrimary source
Estonian Government (RIA - State Information System Authority)· EstoniaLarge Language Models (LLM) · Retrieval-Augmented Generation (RAG)
Government & Public SectorGenerative AIUnknown

City of Helsinki pilots watsonx.ai for multilingual customer care chatbot translation

The City of Helsinki collaborated with IBM Client Engineering and IBM Consulting on a Proof of Value testing whether IBM watsonx.ai could translate foreign-language questions into Finnish for its customer care chatbot, which supports Finnish, English, and Swedish. Watsonx.ai scored 0.90 on the COMET machine-translation metric for English-to-Finnish translation (versus 0.91 for the incumbent system) and 0.87 for Swedish-to-Finnish (versus 0.86), performing on par with the city's existing translation solution.

92/100HighPrimary source
City of Helsinki· FinlandIBM watsonx.ai
Government & Public SectorConversational AIPublic Cloud

Madrid builds VisitMadridGPT on Azure OpenAI to answer tourists in 95+ languages

The City of Madrid, with partner iUrban, built the VisitMadridGPT virtual assistant on Microsoft Azure OpenAI Service to give tourists accurate, real-time, personalized information in more than 95 languages, drawing from the city's official tourism content. Users average 2.8 questions and spend about 4 minutes 18 seconds per session with a low bounce rate, and the city analyzes assistant conversations to identify popular topics and refine its digital marketing strategy.

96/100HighPrimary source
Madrid Destino / Madrid City Council· SpainAzure OpenAI Service

Was this helpful?

Your feedback helps us improve our use case database