AI in Facilities Workshop: Connecting Claude and ChatGPT to Ecotrak

Our AI-powered facilities management suite can automate repetitive tasks, including creating a work order request, summarizing an asset’s repair history, or finding the right service provider for a particular problem. Your team can also use the conversational interface to quickly find answers hidden in large sets of data, analyze spend patterns or asset performance, and even leverage predictive maintenance insights.

Although the possibilities are exciting, a tool is only valuable if it is used correctly. AI interfaces are intuitive, but just like any other skill they require practice to master. Here, we’ll discuss how to write prompts that actually work, and use AI to submit work orders, search assets, troubleshoot equipment, compare vendor proposals, and make better maintenance decisions in seconds.

Connecting Your AI Tools to to Ecotrak

The AI tools you use everyday can now integrate with Ecotrak, through Ecotrak’s MCP server. The setup only takes a few minutes and only needs to be completed one time. The access follows your Ecotrak role — your team only sees what they already can access.

If you need implementation support, reach out to your Ecotrak CSM, and we can support you or your IT team during the setup.

The Most Important Skill: Writing Prompts that Work

As your team begins to use AI-powered tools, one of the most important skills to learn first is how to write the prompts that get you where you want to go. The most important rule is to be as specific as possible in your prompts.

Guidelines for writing a useful prompt:

1. Name the location, asset, or timeframe

2. Say what you want back — a list, a summary, a draft

3. Follow up with the LLM — it remembers your previous conversations

Ecotrak’s AI in Action

Once you have your prompts nailed down, here are tips to keep in mind as you integrate AI functionality into your daily workstreams.

Write in “plain english”

How would you ask one of your facilities management teammates about an asset, work order, location, inventory, or service provider? When you interact with a LLM, search, troubleshoot, and summarize issues using normal language. Just because you’re writing with AI, you don’t need to use a specific format for your questions. The only exception: to ensure you get the most out of your facility data, follow the asset naming convention or location name that you have within the Ecotrak system.

TRY THIS PROMPT: "Show me every open work order for refrigeration equipment at Store #82, with status and assigned vendor."

Create work orders straight in AI

Ecotrak’s platform allows you to create and update work orders without even leaving the chat. You don’t need to open the Ecotrak navigation menu, or switch to an external AI platform. You can use the AI conversation to create a work order with the same designated fields, workflows, and trade assignments as your normal work orders, just with a lot less clicking.

TRY THIS PROMPT "Create a work order: ice machine at Store #1009 isn't making ice. Priority L2, assign our usual refrigeration vendor."

Troubleshoot equipment issues

AI can review an asset’s service history and provide guidance on troubleshooting steps, before you dispatch a tech. To ensure relevancy, there are guardrails on the AI tool about what external sources can be referenced and pulled from (rather than all public data). This feature works best when you have robust service history records within your CMMS platform.

TRY THIS PROMPT: "The walk-in cooler at Store #33 is running warm. Pull its service history and walk me through what to check before I dispatch.”

Summarize anything instantly

You can use AI to boil down months of activity into the most important details. Combine work orders, maintenance history, and technician notes into a narrative that tells you what you need to know before you dispatch or escalate. The tool can also help prompt you what to do next, following your system’s workflows, such as creating a work order or which service provider you might choose. As with other tools, this functionality will improve over time as it learns from your actions and builds on your service history data.

TRY THIS PROMPT: "Summarize the maintenance history and technician notes for asset #5397— what keeps failing and what's been tried?"

Compare proposals and benchmark costs

Use benchmarking for proposals and repair costs, putting vendor proposals side by side and checking them against what you've actually paid for similar repairs. The system can review what proposals vendors have previously submitted to you, as well as recommendations from a regional and national benchmarking standpoint.

TRY THIS PROMPT: "Compare the two proposals on work order #6119364 and benchmark them against what we've paid for similar compressor repairs."

Find your highest-spend assets

Perform maintenance spend analysis faster than ever, even creating visual charts and graphs of the data automatically. You can spot the equipment eating your R&M budget — and decide whether to repair or replace with your own data.

TRY THIS PROMPT: "Which 10 assets drove the most maintenance spend in the last 12 months? Flag any where repair costs are approaching replacement."

Assets and Location AI Insights

Asset Insights are built into the Assets module, without a prompt needed. These insights flag what deserves attention before you open a single record. Here are some examples within the Ecotrak app:

• Completed work orders with an active manufacturer warranty

• Assets within 30 days of warranty expiration

• Assets past useful life or approaching replacement cost

Location Insights are also available within the Locations module, surfacing anomalies and trends by site so you know where to dig in each week. Here are some examples within the Ecotrak app:

• Locations where work isn't getting completed

• Spend trending above the norm for the site

• Open items aging past your response targets

Common Mistakes to Avoid

As your team learns how to integrate AI into their daily workflows, here are a couple common mistakes to avoid:

• Asking without context: Name the location, asset, and timeframe. "My stores" isn't a scope.

• Stopping at the first answer: Follow up. Narrow it, re-sort it, ask why. The second question is usually the good one.

• Skipping the review on AI’s output: Read the draft before a work order goes out — you approve it, AI doesn't.

• Keeping the good prompts to yourself: If a prompt works, share it. One person's shortcut should become the team's routine.

Ecotrak’s AI tools are just like any other skill — you need to practice and play around to understand functionality and opportunity.

To see more examples of how AI can be used in facilities management workflows, watch Ecotrak’s on-demand webinar “AI Workshop for Facilities Teams” here.