top of page
PipeAid

Swipe right on the best match for your pipes

  • Writer: PipeAid
    PipeAid
  • Nov 24, 2025
  • 3 min read

Let’s be honest choosing an AI vendor can feel like online dating. Everyone claims to be “innovative,” “accurate,” and “easy to work with.” But when it comes to sewer inspections, you need more than a pretty pitch deck with baked numbers. You need a partner who gets your pipes, speaks fluent PACP/LACP/MACP, and doesn’t ghost you when things get technical or they end up a gold digger pushing a platform without results.


Here’s your no-nonsense guide to finding the AI vendor that’s actually worth your time (and budget).

 

Step 1: Know What AI Can Do for You 

Traditional sewer inspections are slow, manual, and often inconsistent. AI flips the script by using machine learning and computer vision to analyze inspection footage and spot defects like cracks, roots, and taps… fast and accurately!

 

With the right AI, you get: 

 

  • Speed: No more endless hours of manual review watching dirty videos. Code assist is cute, but it doesn't save you time.

  • Consistency: Every defect gets the same code, every time. 

  • Humans: Yes, humans who understand the AI and ensure you're getting accurate results.

  • Honesty: AI isn't fully autonomous yet...but if you're told 99% AI accuracy, check the ring finger for tan lines.

 

Step 2: Ask the Right Questions

Not all AI is created equal. Here’s what to ask when vetting a vendor: 

 

  • How do you measure accuracy of your AI? Opinions are like armpits...everyone's got two, but we need facts when we're investing in technology to assist our crews. PipeAid, for example, hits 88% accuracy on common defects and flags uncertain ones for human review. If they say they are over 95% accurate, swipe left immediately!

  • How does it fit into your current workflow? You shouldn't have to reinvent the way you work or purchase a heavy platform that does everything you don't need. You are beautiful without all the extras.

  • How do you handle feedback on your AI results? Communication is a two-way street, so if you get one-way answers about "being experts, we're always right" run. Seriously. It's giving red flags.

 

Step 3: Compatibility & Interoperability

Your AI vendor should play nice with your existing systems. That means: 

 

  • No additional cost to export YOUR data: Terms like ‘additional module’ or a ‘license’ are deal breakers. It’s already your data, why are you paying again to access it?

  • Field and office access: Everyone in your organization should be able to use the data, not just those with access to some sewer software platform.

  • No platform, no problem: Inspection data should meet you in the application you’re already using. Love ArcGIS, live in Excel, can’t get enough of Power BI or Tableau… the data should make it to your workbench of choice. 

 

Step 4: Look for ROI 

PipeAid helped one project finish in a year instead of three cutting time and cost by over 50%. That’s the kind of impact you want. 

 

Other benefits include: 

 

  • More footage per day: Crews focus on cleaning, not coding… meaning they’ll inspect more pipe per day to meet inspection goals.

  • Time to value: GIS databases make sewer inspections actionable right away.

  • Staff turnover: Don’t fear promoting staff members to new roles or saying goodbye to old friends, complete coding services lower the barrier for someone new to get started and make an impact.

  • No fatigue, no bias: AI doesn’t get tired or distracted...or can't stop eyeing the chick at the bar... 

 

Also find out:

  • Does my team need additional training?

  • What changes in our workflows from in the field to the back office?

  • What kind of training and support do you offer?

  • What are the costs—upfront and ongoing?

  • How do we access the data from other software’s?

 

Choosing an AI vendor isn’t just a tech decision, it’s a strategic one. The right partner will help you inspect smarter, plan better, and save money. So don’t settle for buzzwords or cheesy pick up lines. Ask the hard questions, demand real results, and find the AI that’s ready to roll up its sleeves and get dirty (digitally, of course).

bottom of page