The News Nobody in the Trades Is Talking About
Last week, PrismML announced Bonsai — a 27 billion parameter AI model that runs entirely on a smartphone. No cloud. No API calls. No internet required.
The tech blogs are covering this as a consumer curiosity. Developers are benchmarking it against cloud models. But almost nobody is asking the question that matters for small business owners:
What does this mean for the person working in a basement with no signal?
Why On-Device AI Changes Everything for Field Work
If you're a contractor, HVAC technician, electrician, or anyone who spends their day moving between job sites, you know the problem:
- Your phone loses signal in basements, backyards, and rural properties
- Cloud-based AI tools become useless the moment you're offline
- You can't look up specs, calculate materials, or draft client messages when you need them most
Bonsai 27B changes this. For the first time, you have a capable AI assistant that:
- Works offline — runs entirely on your phone, no internet needed
- Respects privacy — your client data, photos, and notes never leave your device
- Has no API costs — no per-request fees, no monthly subscriptions for usage
- Is actually capable — 27B parameters is enough for real tasks: reading documents, drafting messages, analyzing photos, answering technical questions
What Field Workers Can Actually Do With This
Let's get concrete. Here's what on-device AI enables that wasn't practical before:
1. Instant Spec Lookup in Dead Zones
You're under a house. No signal. Client asks if the existing wiring can handle a new circuit.
Instead of guessing or making a second trip, you can:
- Take a photo of the electrical panel
- Ask the AI to identify the breaker capacity and wire gauge
- Get an answer in seconds, on-site
2. Real-Time Estimate Adjustments
You discover unexpected work mid-job. Instead of "I'll send you an updated quote later," you can:
- Dictate the additional work to your phone
- Have the AI draft a professional change order
- Send it to the client before you leave the property
3. Documentation That Doesn't Require Typing
Your hands are dirty. You're on a ladder. You need to log what you did.
Voice-to-text with on-device AI means:
- Dictate your work notes hands-free
- AI formats them into professional service reports
- Everything stays on your device until you choose to sync
4. Training Junior Techs Without Pulling Them Off Jobs
You've got the knowledge. Your apprentice doesn't yet.
Instead of stopping work to explain everything:
- Junior techs can ask the AI for guidance on standard procedures
- You review the AI's suggestions before they're applied
- Everyone learns faster without slowing down the job
The Privacy Advantage Nobody Talks About
Here's something that should matter more than it does: when AI runs on your device, your data stays yours.
Cloud-based AI means:
- Client addresses, photos, and notes go to a third-party server
- You're subject to their data retention policies
- A breach at the AI company exposes your client information
On-device AI means:
- Client data never leaves your phone
- You control what gets synced and where
- No third party has access to your business information
For contractors working in homes, handling security systems, or dealing with insurance documentation, this isn't just a nice-to-have. It's a liability reduction.
The Cost Math
Let's talk numbers, because this is where on-device AI gets interesting.
Cloud AI costs (typical):
- $0.50–$2.00 per complex query
- 100 queries/day = $50–$200/day
- Monthly: $1,500–$6,000 for heavy users
On-device AI costs:
- One-time phone purchase (if needed): $800–$1,200
- Electricity to charge: negligible
- Monthly: $0
If you're using AI more than a few times a day, on-device pays for itself in weeks.
What You Can't Do (Yet)
Let's be realistic about limitations:
- Battery drain — Running a 27B model locally uses power. Expect shorter battery life on heavy-use days.
- Speed — On-device is slower than cloud. Responses take seconds, not milliseconds. For most field tasks, this is fine.
- Model updates — You need to download new model versions. Cloud AI updates automatically.
- Storage — A 27B model takes several gigabytes. Make sure your phone has space.
None of these are dealbreakers for field work. But they're worth knowing.
How to Get Started
If you want to try on-device AI for field work:
- Check your phone — Bonsai 27B runs on recent iPhones and high-end Android devices. If your phone is 2+ years old, you may need an upgrade.
- Download the app — PrismML's Bonsai app is available on iOS and Android. (Search "Bonsai AI" in your app store.)
- Test it offline — Before relying on it, try using it in airplane mode at a job site. Make sure it works where you need it.
- Start small — Use it for one task first: drafting client messages, or looking up specs, or documenting work. Get comfortable before expanding.
The Bottom Line
For office-based businesses, on-device AI is a privacy play. For field workers, it's something bigger:
It's the first AI tool that works where you actually work.
Cloud AI assumes you're at a desk with good internet. On-device AI assumes you're under a sink, on a roof, or in a mechanical room with no signal.
That's not just a technical difference. It's the difference between a tool you can use and a tool you can't.
Want to see what else is possible with practical AI automation for small businesses? Check out the SMB Scaleup store for templates and guides built for real-world use.
Have questions about on-device AI for your specific trade? Drop a comment — happy to help figure out if this fits your workflow.
Top comments (0)