What’s Happening With Your SIMs? Soracom Query Has the Answers
Written by: Ryan Carlson
Published: August 29, 2025
TL;DR
- From early access to general release: Soracom Query is now available to all business plan users, bringing AI-powered reporting, monitoring, and BI directly into the Soracom console.
- Natural language analytics: Query transforms raw IoT connectivity and device data into actionable insights with built-in NLP and SQL assistance, no data science skills required.
- Faster problem-solving at scale: From diagnosing unstable SIMs to visualizing fleet performance, Query helps teams reduce engineering overhead, improve operational efficiency, and integrate insights into their BI tools.
Soracom Query, unveiled at Discovery 2025, is now moving from early access to general release. It is a powerful tool for reporting, monitoring, and business intelligence (BI) across an IoT ecosystem of connected devices. From real-time session diagnostics to natural language-based fleet analytics, Soracom Query gives you the visibility you need to operate at scale.
The purpose behind Soracom Query is to unlock your own data and apply it in problem-solving
Modern IoT deployments generate massive volumes of connectivity and sensor data—but analyzing that data has historically required technical teams, manual exports, and significant engineering overhead. Soracom Query was built to solve this problem by making it simple for users across product, support, and operations teams to access and act on insights in real time.
Soracom Query is fundamentally designed to address the challenges customers face in accessing and analyzing their own connectivity data, including session information stored on the Soracom platform and user data sent from devices to Harvest Data. Previously, identifying problematic SIMs, understanding communication patterns, or performing detailed data analysis often necessitated direct contact with Soracom’s support or engineers, leading to additional costs and effort for data export, conversion, and manual processing.
The core problems Soracom Query aims to solve include:
- Reducing inquiries to internal support and engineering teams, thereby enhancing overall business efficiency.
- Allowing customers to analyze and visualize their own platform and user data freely.
- Simplifying the process of identifying problematic SIMs based on communication volume, last session creation time, or other metrics.
How it Works
Soracom Query gives customers direct access to their device behavior and connectivity data through a simple interface designed to minimize complexity and maximize utility. Whether you’re troubleshooting one device or managing 100,000, the system adapts to your needs.
Query can analyze the following data types:
| Category | Description |
| Billing | Monthly billing history, including amounts, currency, and detailed plan information. |
| Connectivity | Network connection details, including country, network, cell tower, connection history, and data usage. |
| User Data | Device data uploaded to Harvest Data time-series data (e.g.: temperature and humidity) and metadata of files uploaded Harvest Files (file lists, sizes, types). |
There are multiple ways of initiating Query requests:
Users can interact with Soracom Query directly through the Soracom Web Console, initiating queries and analyses. For automated processes, API integration is also supported.
Natural Language Processing (NLP) is what makes Query so powerful:
A key feature is its ability to accept queries in natural language, similar to a support bot, removing the need for users to write or execute complex SQL statements. The goal is to enable customers to freely access, analyze, and visualize their own connectivity and device data, such as session history. This capability aims to reduce the need for customers to contact support for data analysis. It allows for analysis using natural language questions, with automatic visualization of results from a Snowflake data warehouse.
You don’t need to be a data-analyst with the built-in SQL Assistant Support:
The AI SQL Assistant will process the request and check for unsupported or incorrect expressions. Like most modern coding agents, the SQL Assistant will suggest alternatives to achieve the desired goal of the initial query.
Data Backend and Visualization
Soracom Query is powered by enterprise-grade data infrastructure. At its core is a Snowflake data warehouse that aggregates connectivity metadata, session logs, and optional sensor payloads from other Soracom services. This foundation allows for both granular filtering and high-level aggregation, providing dynamic visual outputs – from time series to geospatial maps – based on your query intent.
You can direct the service to use a particular visualization or let it automatically choose an appropriate visualization for returned data, such as bar charts for usage stats or maps for location data. It’s a full BI pipeline built into your connectivity console.
Example Scenario: From Natural Language Request to Actionable Insight
Assume we’re managing a fleet of connected delivery vehicles. Unexpected data gaps can raise serious concerns, especially if those vehicles are transporting sensitive goods like refrigerated food or pharmaceuticals. After noticing irregularities in location tracking and temperature reporting, a fleet manager suspects connectivity issues tied to carrier or tower switching along specific routes.
To investigate, let’s ask Soracom Query:
“Show me all SIMs that have switched between different cell IDs or carriers at least 5 times in the last 30 days, and display their most recent cell locations on a map to help detect unnecessary handovers.”
This request touches on multiple analytical dimensions, including session behavior, handover frequency, and geolocation mapping. Initially, the system encountered a compilation error caused by improper use of JSON path expressions and type casting. The assistant automatically identified the problem:
“The previous query contained unparseable SQL syntax, likely due to the use of unsupported JSON path expressions or incorrect type casting. I will simplify the query and ensure all JSON extraction and type casting are valid for Snowflake. Retrying with corrected syntax.”
After retrying and encountering further errors, the assistant corrected the approach:
“Replaced TRY_CAST with the correct Snowflake syntax for extracting JSON values from VARIANT using the :: operator (e.g., CELL:ci::STRING). This should resolve the SQL compilation error.”
The assistant successfully:
- Extracted cell switch events from session metadata
- Counted switches per SIM
- Filtered for devices with frequent transitions
- Mapped the most recent locations of these SIMs
Upon success, the system provided additional refinement tips:
“The data looks good and the map visualization is appropriate for showing the most recent cell locations of SIMs with frequent handovers. Note that there are duplicate rows for the same SIM and cell location; if you want to remove duplicates, you can add SELECT DISTINCT to the final query.”
Visuals can be viewed on a map, as a table of information, or downloaded as a .csv file.
This example highlights the assistant’s ability to self-diagnose issues, revise SQL logic, and deliver production-grade results even for users without SQL expertise.
This would help the fleet operations support team in the above scenario:
- Identify SIMs with excessive switching, which is often a sign of unstable or “ping-ponging” connections.
- Visualize affected devices geographically, so they can pinpoint high-risk areas like mountain corridors or urban fringes.
- Cross-reference problematic regions with known dead zones or misbehaving towers.
- Make informed changes, such as:
- Adjusting modem firmware carrier preferences,
- Configuring static PLMN selections,
- Switching to a different IMSI or coverage profile for better regional performance.
- Adjusting modem firmware carrier preferences,
Integration Capabilities
Soracom Query is not just an internal dashboard, it’s designed to integrate with the broader analytics ecosystem.
- Enterprise users are provided with a read-only account for the Snowflake data warehouse, allowing seamless integration with BI tools such as Tableau or Looker Studio.
- Soracom is also exploring options to visualize query results within Lagoon (Soracom’s monitoring dashboard) or to utilize them via Flux, enabling automated workflows and alerts.
- While Soracom Query does not strictly depend on services like Soracom Air or Harvest Data/Files, valuable query results are best obtained when these services are in use.
This flexibility makes Soracom Query a cornerstone of any scalable, intelligent IoT deployment.
Key Benefits and Use Cases
Soracom Query isn’t just about accessing data, it’s about turning insights into action. From field diagnostics to executive-level reporting, it delivers results that help teams work smarter and faster.
- Enhanced Operational Efficiency: Detect issues like weak or unstable signals, frequent carrier switching, or session drops quickly, often before customers notice them.
- Proactive Problem Identification: Identify SIMs exhibiting abnormal behavior over a given time frame and flag them for review.
- Geographical Insights: Visualize device connectivity locations based on cell tower data, helping with RF optimization, fleet tracking, or audit compliance.
- Automated Alerting: Integrate with Flux to trigger alerts when threshold-based anomalies are detected in your sensor data.
- Simplified Analytics: Use natural language queries and receive easy-to-digest statistical summaries without the need for an SQL.
- Reduced Development Overhead: The low-code/no-code design means less time building dashboards or exporting data, and more time acting on insights.
Built-in Example Queries in Soracom Query
To help teams get started quickly, Soracom Query includes a set of predefined query examples that address common operational and analytical needs. These built-in examples not only return data but often include visualizations like tables, time charts, or maps, reducing the need for post-processing.
Here are several examples available in the console:
| Built-in Example Query | What it Does | Natural Language Equivalent |
| Top 3 SIMs with high event counts | Lists SIMs with the highest session activity. | “Which SIMs have the most session events this week?” |
| Monthly billing total | Aggregates billing by month. | “What are my total monthly data charges?” |
| Real-time Device Location Tracking | Maps current SIM locations using tower data. | “Where are my devices right now?” |
| SIM count by status | Breaks down SIMs by active/inactive status. | “How many SIMs are active or suspended?” |
| SIMs with high data usage | Identifies top 5 SIMs by usage. | “Which SIMs used the most data this month?” |
| Roaming SIMs by country | Shows roaming distribution by country. | “Where are my roaming SIMs located?” |
| SIMs connected to specific towers | Filters by cell ID or region. | “Which SIMs are using tower ID X?” |
| Inactive SIM list | Returns SIMs with no recent activity. | “List all inactive SIMs in the last 30 days.” |
| Data transmission pattern | Analyzes usage by hour/day. | “What time of day is most data being sent?” |
| SIM distribution by network | Breaks down carrier usage. | “How many SIMs are using each carrier?” |
Each of these can be run as-is, modified for your deployment, or embedded in dashboards or automated workflows. They serve as a helpful starting point for users new to the platform and as rapid diagnostics for experienced operators.
Technical and Security Considerations
Soracom Query is engineered with reliability, transparency, and governance in mind.
- Data Sources: Connectivity and session data are pulled from Soracom’s trusted data lakes in Snowflake and BigQuery.
- CRE Support: Soracom’s Customer Reliability Engineers (CREs) use the same toolset to troubleshoot customer deployments, ensuring fast and informed support.
- Security Controls: To prevent misuse, Soracom Query includes limits on execution time and result sets. Usage-based billing ensures responsible query behavior while preserving accessibility.
Pricing and Availability
Soracom Query is available under the Business Plan, making it accessible to customers ready to scale operations with data-driven intelligence.
- A two-month limited trial is available for qualifying users.
- The standard pricing model includes 2,000 queries per month, with additional tiers and custom enterprise pricing available.
- Japanese customers can access pricing directly in-console, while US and UK customers are encouraged to contact the Soracom sales team for details and onboarding.
Looking Ahead
Soracom is actively gathering feedback to expand the number of predefined queries, improve natural language understanding, and deepen integration with other platform services like Lagoon and Flux. The roadmap includes custom alert logic, scheduled reports, and enhanced query templates.If you’re ready to put your IoT data to work without writing a single line of code, Soracom Query is ready to help.
………………
Got a question for Soracom? Whether you’re an existing customer, interested in learning more about our product and services, or want to learn about our Partner program – we’d love to hear from you!
