AI systems and governance
AI built with intention.
I design and build AI systems around how your business works, then help your team put them to use. Custom tools, workflow automation, and oversight for small and mid-sized businesses.
Free, 20–30 minutes. Bring a task you want to improve.
- Raw drone footagebecomesautomated video production with script, AI voiceover, and captions
- Chronological imagesbecomescomputer vision extracts data for charts
- Resale listing workflowbecomes75% less time per listing
- AI governancebecomesdocumented review policies and accountability
- A month of AI outputbecomesa review of errors in a sample of its work
Services
What I can help you build.
Workflow assessment and planning
We look at how a task gets done, where time goes, and what information is available. You get a recommendation on what to automate and a scope for the build.
Explore workflow assessment and planningCustom AI systems and applications
I design and build tools around a specific business job, from an internal assistant to an application your customers use. We agree on what it needs to do before development starts.
Explore custom ai systems and applicationsWorkflow automation
I connect the steps that otherwise require repeated manual work. That can mean producing a video from raw footage or getting an item from intake to a published listing.
Explore workflow automationKnowledge and data tools
I build tools that find information in your documents, extract readings from images, and turn the results into usable records and charts.
Explore knowledge and data toolsAI adoption and training
I work with the people who will use the system, test it against their tasks, and document how it works. Training covers everyday use and how to recognize a result that needs review.
Explore ai adoption and trainingAI governance and ongoing support
We define who checks the output and how problems are handled. After launch, I can maintain the system, review samples of its work, and make changes as your needs develop.
Explore ai governance and ongoing support
Work
Selected work.
Resale operation · Process improvement
80 → 20 min
75% less time per listing
I rebuilt the intake-to-listing workflow around an AI assistant the business already used. Photo requirements and cited sales comparisons replaced open-ended back-and-forth. Listing time dropped from 80 minutes to 20, with no new software purchased.
Read the project: 75% less time per listingCavlon · Video automation
Raw footage to a scripted, voiced video
I built a system that automates video production from raw drone footage, including the script, AI voiceover, and captions. Each YouTube Short goes to a person for review before posting.
Built. Further testing is awaiting additional footage from Cavlon.
Read the project: Raw footage to a scripted, voiced videoCREATIVE at ASU · AI application
Resume and portfolio feedback in Discord
I built and maintained an AI bot for a 100+ member student community. It reviewed uploaded resumes and used screenshots to evaluate live portfolios. I kept it running through deployment failures and model changes from January through August 2026.
Read the project: Resume and portfolio feedback in Discord
More projects: computer vision, proposal drafting, and a member portal
Cavlon · Computer vision
From chronological images to capacity charts
I built a tool that extracts battery test readings from images, orders the readings over time, and charts battery capacity across a full discharge. It turns a sequence of screen photos into documented test results.
Read the project: From chronological images to capacity chartsFederal contractor · Graduate capstone
Proposal drafting with source material and review
I led a four-person team and built a working prototype that retrieves internal information to support proposal drafting. Writers review the retrieved material before it enters a draft. The handoff included an adoption plan and documentation.
Academic prototype. Business improvements were projected, not measured in production.
Read the project: Proposal drafting with source material and reviewArizona Spurs · Web application
A website with a working member portal
I rebuilt the club’s website and added member sign-in, role-based access to documents, and a match calendar. The work included requirements, acceptance criteria, and accessibility checks across the site.
Read the project: A website with a working member portal

Chandra Carr · Founder & AI Implementation Consultant
Built with the people doing the work.
Before founding OPZET, I ran a small business and supervised a manufacturing shipping and receiving team. I have been responsible for the work a new system is supposed to improve.
At Times Microwave, I relaunched an inventory system the site had never adopted, learned it from corporate headquarters, and trained the eight-person team. I also built an Excel-based planning system that went into daily use.
My BS in Human Systems Engineering and MS in AI in Business from Arizona State University bring those operations and technology experiences together.
Speaking and workshops
How to Get Your AI to Stop Lying to You
A hands-on session for small business owners at The AI Exchange, Workuity, in November 2026. We’ll use code to count the data and AI to explain the results, then practice spotting made-up numbers.
I also led AI workshops at ASU and helped develop AI usage guidelines and evaluation criteria for more than 100 students.
Ask about a workshopWorking together
Start with one workflow.
01 · Define the job
Assess and scope
We start with one business problem, map the current process, and establish a baseline. You get a written recommendation, acceptance criteria, and a fixed price for the build.
02 · Put it to work
Build and launch
I build and test the system with you, then document it and train your team. You see progress along the way and review the work against the criteria we agreed on.
03 · Keep it working
Support and oversight
Ongoing maintenance, output reviews, and improvements based on actual use. I can also assess and support an AI workflow your business already has.
Scope, timing, and pricing are agreed in writing before paid work starts. Monthly support is a separate engagement.
Discuss your workflowGovernance
Who checks the AI’s work?
An AI answer can read well and still be wrong. Every tool I build includes a process for reviewing its work and recording mistakes, so your team knows what needs attention.
- 01
Agree on the checks
Define what the tool needs to get right, what counts as a mistake, and who decides whether the work is acceptable.
- 02
Test the tool
Use representative inputs and compare the results against the acceptance criteria before launch.
- 03
Set review responsibilities
Decide which outputs need a person’s approval and who handles that review.
- 04
Record mistakes
Document how to report a problem, keep the relevant output, and get it corrected.
- 05
Plan ongoing reviews
If you choose ongoing support, agree on a review schedule and sample outputs for errors that went unreported.
Ready to stop exploring AI and start building with it?
What automation would make your life easier?
Bring it to a free 20 to 30 minute call. We’ll talk through whether AI could help and what it would take to build the tool.
Book a callPrefer to write first? Send me a message.