In Part 1 of this series we demonstrated how to leverage AI for faster, more intuitive ways to access real-time insights. Part 2 demonstrated how to extend the core components used in Part 1 to add an embedded Chatbot interface into Losant. In Part 3 we extend those concepts even further, adding the ability make data predictions based on previous behavior.

This proof-of-concept combines four components:

• Chatbot UI as the conversational interface
• Skylytics’ MCP server to interpret IoT-related requests
• Losant’s IoT platform as the data source
• Neural Network to make predictions

The result: the ability to predict future values and associated alarms for initiatives such as preventative maintenance.

‍

How the Demo Works

1. The user starts with a plain-English request within a Losant App UI.
2. For example: “What will thetemperature values for the  Workspace multi-sensor be at this time tomorrow”
3. Claude passes the request to Skylytics’ MCP server.
4. The MCP layer handles device discovery, mapping conversational intent to actual IoT resources within Losant.
5. For predictions the MCP calls a Neural Network to make a prediction.
   • The Neural Network retrieves a set of previous data values for the given sensor.
   • The Neural Network makes a prediction based on the data.
6. The system then checks to see if any alarm conditions will be fired based on the predicted value and if so informs the user.
7. Claude presents the result back to the user.‍
   
   • Automatically renders a dashboard displaying real-time readings.
8. No manual lookups, no dashboard hunting, no queries—just answers.

This workflow demonstrates a fundamental shift: conversational interfaces are no longer “nice to have” add-ons. With the right integration strategy, they can become the front door to your entire IoT ecosystem.

Why This Matters for IoT-Driven Organizations

Most organizations struggle with three things:

(1) finding the right sensor data when they need it, and

(2) turning that data into something visual and actionable.

(3) Preventing unplanned downtime

By layering Neural Networks along with conversational AI and a robust platform like Losant, Skylytics eliminates these barriers. This approach offers:

• Hands-free access to IoT data — ideal for field techs, facilities teams, and operators
• Real-time dashboards built on demand
• Reduced friction in troubleshooting, monitoring, and analysis
• A more accessible IoT experience for non-technical users
• Advanced notice of any potential unplanned downtime or issues
  

It’s a glimpse into how AI-driven automation and IoT platforms are converging—and how organizations can use both to accelerate digital transformation.

Skylytics + AIOT: Turning IoT Data Into Decisions

Skylytics has been delivering IoT-based solutions for years, including fully managed IoT deployments, sensor integrations, workflow automation, data pipelines, and dashboard development.

This latest demo underscores our commitment to helping clients go beyond traditional dashboards—unlocking new ways to interact with devices, understand their environments, and act quickly.

If your team is:

• exploring conversational interfaces for IoT,
• already using IoT and want to extend its capabilities, or
• looking to modernize the way you interact with your connected devices…

We’d love to help you make it real.

Ready to See What’s Possible?

Whether you want to replicate this demo or build your own AI-assisted IoT workflows, Skylytics can help.

Contact Us

‍

Read Part 1 again.

Read Part 2 again.

‍

‍