# BADDA: Turns acquisition data rooms into structured, reviewable due diligence

> BADDA extracts financials and operating data from deal documents, flags anomalies and builds diligence reports for analysts to review.

Source: https://co-worq.com/built-by-coworq/badda

Built by CoWorQ · Business acquisitions

# BADDA

Turns acquisition data rooms into structured, reviewable due diligence.

BADDA extracts financials and operating data from deal documents, flags anomalies and builds diligence reports for analysts to review.

CoWorQ built BADDA for business acquisitions. It follows the same governed pattern as QRION, and it is ready to deploy today. Interested? Talk to us and we’ll set it up for your business.

[Talk to us about BADDA](https://calendar.app.google/yF5awaB5GSSMDumE8)[Watch the demo](#demo)

## What is BADDA?

BADDA is an acquisition due diligence workflow that CoWorQ built for buyers of senior living communities, nursing homes, strip malls, gas stations and retail businesses. Agent steps read financial and operating documents in PDF, Excel or scanned form, extract structured data, and cross-check it across documents and periods. Anomalies are flagged for analysts, who can ask questions in plain language. Analysts review the findings before BADDA produces pro formas, compliance checklists and diligence reports, and every extraction is recorded.

See it in action

## BADDA demo

BADDA demo.Uploading deal files and importing a Box folder, reviewing flagged anomalies, and exploring the extracted financials.

Transcript

Once you log in, this is the view you will see. I'm going to drag four files onto this and do an upload and analysis. You notice that it's going to process those four files, and it immediately invokes three workers, distributing three of those files to each.

Those workers read the files, analyze them, and convert them into the back end database format.

Since the workers are processing them independently, it will take about one third as long to process them sequentially.

There it finished. One of the files is remaining. Now for each file, if any anomalies are found in the data, they will appear in this section.

It shows, for example, that some details in the CLE trend file are incorrect. You can accept it as is, exclude it from the file, or mark it for review.

There are two ways to do this. One is from the individual selection on the right or use the bulk actions at the top. I am going to accept each of them as is for demo purposes.

Now I'm going to show you how it connects to the box or Google Drive and processes files. Here, I'm going to connect to the box by logging in and authorizing this application to access your box drive. This login is whatever you have set up on the Box side.

I have enabled MFA.

Have to enter the MFA code from the authenticator app. Once it submits, it asks you to authorize the BADDA application to access your files. You can now see the file explorer, where you can navigate to any of the directories in your Box drive. I'm going to go to the top of the folder and select all the files. It read through all of the subdirectories and picked up 125 files.

And I say import. It sets up back end workers to access these files.

Now you see it's downloading those 125 files from Box onto the server to analyze each file. Now it's going to process it, analyze all 125, and see which are ready for import.

While it's doing it, let's look at the property explorer from the previous four files we imported. Here, you see the CLE, DLA, PCO, and SAR.

For every one of those nodes, you can see information, the financial summary, and so on, and you can drill down further for each year. It will show the summary of the financial data.

Further for each year, you can see by the different units, data, IL, AL, and MC.

And further down, you can see the kind of breakdown for rent, the net income.

This is the trend for the different data points. You can move the group nodes to rearrange it the way you like. Now let's look at some query. Here is where you can enter natural English language text. You see some examples here. You can query, and it will go through, understand what your question is, and run a query against the database and come back with the data.

Here it shows what it found in in plain English text. You can see what is the data that is behind this, or you can see the chart of that data plotted. Depending upon the data, the plot may be

a bar graph or a line graph. Now let's look at the reports. You can select properties and say generate report.

If you select more than one property, it will put into a ZIP file. Each property has one report, and that's how you will download it. Let's go back and see how the box upload worked, and there it is still processing.

You can see that it's allocated three workers. There are five financial files out of the total 125 files and one of the file it processed. So it has the next one available.

There is one file it rejected, and the reason is it's got a bad format.

Now again, each file it processed, it comes up with any exceptions that you have to make a choice.

I am accepting as is for demo. You may have a different reason to look at that and make some changes. And there it has finished all of the file processing. This concludes this demo.

The problem

## Where business acquisitions work gets stuck

- ### Weeks of manual document review

    Rent rolls, P&Ls, cost reports, leases and surveys arrive in every format. Analysts rekey them before any real analysis starts.

- ### Missed anomalies

    Inconsistencies between documents, such as reported occupancy that does not match revenue, are easy to miss when each file is read on its own.

- ### No standard process across deal types

    A nursing home, a strip mall and a gas station each need different checks, so every deal team builds its own spreadsheets.


The workflow

## Agent steps inside a governed process

Agents prepare the work. People decide where judgment matters. Every step is recorded.

1.  1

    System step

    ### Documents ingested

    Financials, rent rolls, leases, surveys and reports are uploaded or pulled from Google Drive, Box or SharePoint.

    Role

    Intake

2.  2

    Agent step

    ### Extract and normalize

    Agent steps extract line items into a common structure across different management company formats.

    Role

    Extraction agent step

3.  3

    Agent step

    ### Cross-check and flag

    Data is compared across documents and time periods, and inconsistencies or unusual patterns are flagged.

    Role

    Analysis agent step

4.  Approval gate

    ### Analyst review

    An analyst reviews flagged items, asks follow-up questions and approves the data set for modeling.

    Role

    Deal analyst

5.  5

    Agent step

    ### Build models and reports

    Pro formas, DSCR models, compliance checklists and diligence summaries are generated from the approved data.

    Role

    Reporting agent step

6.  6

    System step

    ### Record and export

    Results are stored for the portfolio with a full audit trail and exported to Excel, CSV or JSON.

    Role

    Record


Capabilities

## What BADDA does

- ### Multi-format extraction

    Reads PDFs, spreadsheets and scanned images and extracts structured financial and operating data.

- ### Vertical-specific checks

    Cost reports for nursing homes, lease abstracts for retail centers, fuel volumes and environmental reports for gas stations.

- ### Anomaly flagging

    Cross-references documents and periods and flags inconsistencies for an analyst to resolve.

- ### Plain-language questions

    Analysts ask questions about the data and get charts and narrative answers.

- ### Portfolio data store

    Extracted data stays in one place and can be compared across properties in a portfolio.

- ### Audit trail and roles

    Every extraction, user action and AI call is logged, with admin, manager, analyst and viewer roles.


The reusable pattern

## Same shape, any process

BADDA shows the QRION extract-reconcile-review pattern for large document sets, with analysts approving data before it feeds models. It applies to lender packages, vendor onboarding and audit preparation.

[How QRION runs any process](https://co-worq.com/platform/qrion)

- [LUPA

    Mortgage lending

    ](https://co-worq.com/built-by-coworq/lupa)
- [SafetyScout

    Construction safety

    ](https://co-worq.com/built-by-coworq/safetyscout)
- [KeyWarden

    Property management

    ](https://co-worq.com/built-by-coworq/keywarden)
- [SentriCash

    Cash advance and consumer finance

    ](https://co-worq.com/built-by-coworq/sentricash)

## Want BADDA for your business?

It is ready to deploy today. Talk to us and we’ll set it up for your business.

[Talk to us](https://calendar.app.google/yF5awaB5GSSMDumE8)[Contact us](https://co-worq.com/company/contact)