top of page

Buddy's Bark Tank

Updated: Oct 20

One of my favorite things we do at PowerApps911 is our annual internal competition, Buddy’s Bark Tank, where teams build creative solutions to real company challenges.


AI has shown up in past projects, but this year we made it a requirement. Why? These projects move the company forward, strengthen our team, and fuel learning. We are ALWAYS learning over here, and the best way to do that is by building something real. You push yourself to figure out the solution, and we’d rather our consultants take those first tries on internal projects, not your production app. And boy, did we learn some things!


The Bark Tank Hackathon by PowerApps911

First, a little about each of the five projects:


Shane’s YouTube Agent

This tool was built to solve a very common question: “Which of Shane’s YouTube videos talks about _____?” This team built a Copilot Studio Agent that accesses the transcripts of ALL of Shane’s YouTube videos. And since it’s AI, you don’t need to worry about exact keyword matching. No more posts on Teams with “Hey, does anyone remember which video Shane talks about cascading drop-downs?” Nope, just go to the Copilot Agent, type your question, and get an answer! This solution is being used internally now, but we can see a few people being interested in it. 😊Check out all the tech used in it!


Shane's YouTube Agent Breakdown

Project Task Tracking

Organizing tasks, change requests, and bugs on projects with external clients can be time-consuming and messy. This tool fixes that, and not only does it fix it, but AI also analyzes the meeting transcripts and suggests tasks based on items discussed in the meeting. The project manager reviews the suggestions and adds them to the project task board. Tasks can also be added manually by the project team, and customers have external access to see status and submit tasks, bugs, and change requests. We need to implement a few features and do some internal testing, but we cannot wait to start using this with our customers!


Project Task Tracking solution

Quoting

Oh man, quoting is a task that in all my years in consulting, I’ve never heard anyone excited to do. This team decided to tackle some of the challenges. Using machine learning, this solution extracted build details from previous statements of work (unstructured data), compared estimates to actuals in our timesheets, and then used that data to generate a quote for new work. It even creates a draft of the statement of work based on that data. All consulting companies need something like this! Here is a sample of the setup for the simple statement of work output:


Quoting solution

Marketing Data Fabric Project

One team used Fabric, Fabric Agents, and machine learning to bring together data from all over our company to make better marketing decisions. This project pulls data from our CRM, QuickBooks, timesheets, ActiveCampaign, LinkedIn, Thinkific, and more into a single Lakehouse to help us make better decisions. The Fabric Agent makes it easy to ask questions and get insights from all that data using natural language, turning complex analytics into quick, actionable answers.


Marketing Data Fabric Project solution


Buddy’s Wag Wallet 

We love what we do, and we want to celebrate that. Buddy’s Wag Wallet is a solution for company praise. Points can be awarded and redeemed in the company rewards center. There’s even a fun AI tool that uses image generation and modification to create and modify a Buddy image the team can show off in the app. This tool was built using Dataverse, Model-Driven App, Canvas Apps, Power Automate,  Power BI, and AI. It’s not quite ready for release, but it was a fun showcase of what’s possible.

Buddy’s Wag Wallet solution

These projects all have big impact potential, and even if they never make it to production, the lessons learned were invaluable. Of course, the team learned a bunch of new nuances to building in these tools, but even more importantly, they started to demystify the costs associated with these tools. That learning led us to put guardrails in our own Microsoft 365 environment, the same kind of guardrails you might want to reach out to us about setting up in yours. You know, so you don’t use 1.5 million AI Builder credits when your subscription is for 20,000.


We love sharing what our team learns because it often sparks great ideas for others. Contact Us and let us know what it inspired for you!

 
 
 

Comments


bottom of page