Making sense of where your money is going with machine learning

All organizations need to “keep the lights on,” but as they grow and operate, the effort to keep lights on turns into an effort to keep energy spending under control. Add multiple facilities to the mix, and there’s a continuous flood of utility bills from multiple energy suppliers using unique forms, rates, and layouts.

This is where you step in.

The goal is to calculate the total cost across multiple dimensions and observe the impact that different factors like changes in hardware, weather, and occupancy have on consumption. You’ll want to ensure that an organization’s utility spend is aligned with their actual energy needs. With Impira AutoML, you can quickly and effortlessly spot trends in utility expenditures, find outliers, and make sense of an assorted (and growing) stack of energy bills. Harness the power of automation exactly in the way your business needs it.

With changes on the horizon, to be informed is to be prepared for anything

As we all prepare to navigate a hopefully soon-to-be post-COVID world, businesses will need to adapt to a resurgence in human activity, as well as adapt to the emerging renewable energy market that continues to gain momentum in 2021. As Enel X reports, the energy market landscape is on the cusp of monumental shift as renewables become increasingly more cost competitive and many state legislatures mandate transition. These movements are reflective of the market as demand by customers and investors for carbon-neutrality and emissions reductions continue to mount.

As an energy or utility auditor, you can help your clients be proactive and be prepared for major changes such as these. Impira builds on existing character recognition technology and goes further by letting you interact with the results and make changes, giving you more freedom and control over your data. By utilizing Impira AutoML, you’ll be able to make better decisions faster and your consultation can incur more meaningful change for any organization through the power of organized data. We’re all in for some major changes, so why not take the opportunity to pick up the pace and get ahead?

Going lightspeed with OCR + Impira AutoML

Simplicity and speed is at the core of what we can offer. With Impira AutoML, teams can perform better by getting more work done in less time. This results in not only cost savings, but also higher revenue and opportunities to outpace your competitors in presenting quotes and estimates.

Let’s give it a go using the sample order form packet available to download at the top of this page.

How to set up Impira AutoML to manage utility bills

Step 1: Sign up for a free Impira account

Go to impira.com/signup to create your free account. There’s no need for any credit card or personal information to get started, just an email address. Your free account can process up to 200 file units (pages), so you can jump straight into the rest of this tutorial.

Step 2: Create a new collection

Start by creating a collection, a folder that holds and organizes your different file types. Click the “+” symbol in the left-hand sidebar next to the word “collections.”

Give each collection a unique name that describes what types of files you’re uploading, like “Energy Bills.”

‍

Step 3: Upload your sample energy bill files

Click the name of your newly created collection in the left-hand sidebar. When choosing which files to upload into a collection, make sure they all have in common the same kind of data you’re looking to extract (e.g. Account number, Invoice Date, kWh).

One way to upload documents into Impira is to drag and drop your files directly onto the browser window. You also have the option to upload files via Amazon S3 or Dropbox, and you’ll soon be able to upload files directly from your email inbox.

Impira AutoML will get to work immediately to process the data within your documents. As soon as it finishes, you’ll see that each file has its own row in a spreadsheet.

Step 4: Train a unique AutoML model for each collection

Unlock your initial data from your bills

Impira AutoML takes standard optical character recognition to the next level by giving you the freedom to make the changes you need. Training this collection’s machine model to fit your needs is simple. Let’s begin by grabbing the first bit of data you want from each document.  

1. Double-click on the first file to open up the mark-up interface.

2. Let’s start with identifying “Account Number.” On the document, click the number you want to select. Then, on the right-hand sidebar, name this field by typing in “Account Number” and clicking “Add Field.” Close the interface by clicking the “X” in the top-right corner, and see an “Account Number” column was automatically added to your spreadsheet.

Impira AutoML takes this cue and runs with it, immediately working to identify and retrieve account numbers from the rest of the documents in this collection. Your page will begin to fill up with account numbers.

Making corrections and adjustments

Some cells may initially be blank, or have a colored flag displayed on the left-hand side of the cell. This flag color is a visual confidence indicator of Impira AutoML’s prediction: green meaning confident and red meaning not confident. Because Impira AutoML is a continuously learning system, you can boost these confidence levels by confirming correct information or correcting misidentified information. These affirmations increase Impira AutoML’s accuracy by teaching it what you want it to do. The more it learns, the more you’ll see red flags turning green, indicating high confidence.

Double-click a red-flagged cell and take a look to see if Impira AutoML identified the right figure you want. Click the check mark on the right-hand side if it’s correct.

Sometimes, Impira AutoML will misidentify a piece of information (e.g. accidentally highlight a “Previous Balance” figure instead of the “Total Amount Due” figure). Easily correct this by clicking the “X” in the right-hand sidebar and use your mouse to redraw a box around the correct figure.

As you continue to confirm or correct predictions across all your collections, Impira AutoML will automatically update each of their custom machine learning models, making them more and more useful for you.

Retrieve other information from your documents

Repeat Step 4 to keep extracting different information from your documents (e.g. Invoice Date, kWh, Total Amount Due, etc.). Every time you do this, your table will grow and display the information you need to see.

Step 5: Export your data

You’re at the finish line. You’ve pulled out all the data you need and you’ve made all the confirmations and corrections to see all red flags turn green, indicating high confidence. It's time to export your data.

Export via CSV

Export and download this table as a CSV file that you can open in Excel or Google Sheets. Click the “Download” button in the top-right corner and select “All files records (CSV).”