ConnectQuery
Artifical Intelligence - AI for infrastructure GIS data analysis

Query, visualize, and unlock the full potential of your geospatial data with ConnectQuery—available on our secure SaaS platform or directly within your GIS tools.
Automate quality checks, detect anomalies, and secure your data repository with rules and standards tailored to your operational needs.

Écran d'ordinateur avec ConnectQuery en utilisation

Interact with Connie, our AI-powered assistant, using natural language. Instantly get clear, visual, and actionable insights.
Use ConnectQuery on our dedicated platform, directly within your GIS tools, or through our Web GIS platform, ConnectServices.

A new way to leverage geospatial data

With ConnectQuery, there’s no need for complex technical queries or expensive BI tools. This AI for GIS infrastructure data allows you to interact directly with your operational databases using natural language.
Our goal: to democratize access to geospatial analytics and accelerate decision-making, both in the field and in the office.

ConnectQuery provides the infrastructure, security, and visualization engines that bring your operational data to life.

What you can do with ConnectQuery

1.
Visualize geolocated assets: antennas, fiber optics, access points, street furniture…
2.
Explore dynamic maps tailored to your operational needs
3.
Create dashboards automatically from simple questions
4.
Identify priority areas for field operations or investments
5.
Instant access to key metrics: network coverage, equipment rate, availability…

Meet Connie, your AI data copilot

Connie is the conversational interface of ConnectQuery—your AI assistant that simplifies everyday interactions with your data through its key capabilities:

  • Natural understanding: a next-generation LLM translates your operational questions into precise SQL queries.
  • Instant results:interactive maps, KPIs, charts, and tables are displayed in real time, with location-based details (records, attributes, history).
  • Autonomy:no complex setup required—no scripts, no BI training needed.
  • Operational context : Connie understands your domain vocabulary (antennas, fiber routes, access points, street furniture, etc.) for smooth and relevant interactions.

AI built for your geospatial data

ConnectQuery is built to meet the needs of infrastructure managers:

  • Local authorities and technical departments
  • Network operators (energy, telecom, water, transport, etc.)
  • Engineering firms and infrastructure operators
  • Business units, project managers, and public sector decision-makers

Whether they have advanced data skills or not, your users can query the database in natural language and get reliable, actionable insights instantly.

Our vision with ConnectQuery

We developed this AI for GIS infrastructure data to overcome the limitations of traditional tools.
ConnectQuery is built to meet the needs of infrastructure managers:

  • Simplify access to data for everyone
  • Make interactions as natural as a conversation
  • Reduce analysis time to just a few seconds
  • Deliver a new generation of BI: fast, intuitive, and accessible
  • Accelerate decision-making

ConnectQuery represents GenBI—the next generation of Business Intelligence.

Key takeaways

  • AI specialized in GIS infrastructure data
  • A natural language interface, with no training required
  • Instant answers in the form of maps, tables, and metrics
  • Guaranteed security, with no changes to your databases
  • Full autonomy in analysis and decision-making
Question posée par un utilisateurs et réponses de Connie, notre agent IA, sous différents formats (graphique, carte et tableau)

Ready to unlock the full potential of your data ?

ConnectQuery and Connie transform your GIS data into a strategic advantage.

Start today

FAQ – AI for geospatial data exploration

What controls do local authorities have over the generated results? AI can make mistakes or approximations. 

Each generated result goes through an intermediate analysis step before being validated or shared. This mechanism acts as an intelligent filter, checking content consistency, compliance with internal data, and accuracy against authorized sources.

Responses are always built from internally validated documents, using a contextual search engine that retrieves and links relevant information before generation (RAG approach). This ensures full control over sources and guarantees traceability of every piece of data used.

We also provide technical users with a monitoring and audit platform. It allows them to visualize, control, and manually validate the information generated by the AI. Each response can be approved, corrected, or rejected according to the organization’s criteria.

The entire system ensures a transparent, controlled, and secure operation, where AI supports decision-making without ever replacing human oversight.

Do you rely on APIs (such as OpenAI APIs), or are your AI models hosted internally?

We use APIs to connect to large language models (LLMs). As part of the demonstration presented during the webinar, we used GPT-4.1-mini. However, the architecture we have developed is model-agnostic: it enables API-based connections to any LLM, whether hosted locally (e.g., Llama) or accessed remotely (e.g., Gemini, GPT, Mistral, etc.).

Generative AI applied to GIS often involves sensitive data: how do you handle confidentiality and GDPR compliance? 

Our AI solution only works on the structure of the data, not its content. No data is read or stored. Data security and confidentiality are fully preserved. 

How can you generate statistics by feeding the AI only with contextual data?

Through high-quality documentation, where information is explicit rather than implicit. The LLM then transforms this documentation into a query method for a database. This method is applied by our conversational agent to retrieve the data needed to generate the response. As a result, no database content is ever transmitted to the AI.