You’ve been there: back at the truck, squinting at blurry photos, trying to remember the exact condition of a corroded valve or the routing of a wire run. The proposal writing process becomes a puzzle, eating into time better spent on the job. For specialty trade contractors, the key to automating service proposals lies not in complex software, but in how you capture site data. AI can transform site photos and voice notes into structured scopes and material lists, but only if you feed it the right intelligence.
The Principle: Structured Context is Everything
AI needs more than random snapshots; it requires structured context to accurately identify components, assess conditions, and infer scope. Your goal is to methodically capture the what, where, and why of a job. Think of it as building a visual and verbal story for the AI.
The Core Framework: The Four Essential Shots & Narrative Voice
Follow this simple sequence on every site visit:
- The Establishing Shot: Before anything else, take one wide-angle photo of the entire room or area. This is the "big picture" that shows spatial relationships. For a plumbing re-pipe, this is the whole basement ceiling showing the existing pipe runs.
- The Subject Shot: Move in on the specific component needing work—the faulty panel, the leaking joint.
- The Context Shot: Show what’s around the subject. Where does the conduit run? What’s adjacent to the leak? This captures access constraints and connections.
- The Reference Shot: Capture serial plates, measurements, or existing material specs.
This visual story must be paired with a clear audio narrative. Start each voice recording by stating the category, like "Recording: Main Floor Electrical Assessment." Then, dictate using a consistent checklist:
- Item Identification: "Main service panel."
- Current State: "Corrosion on all terminals, 100A capacity."
- Recommended Action: "Replace with new 200A panel."
- Labor Notes: "Install requires new meter base and grounding upgrade."
- Scope Summary: "Full service panel upgrade to 200A."
Mini-Scenario: An electrician finds an undersized panel. They take an establishing shot of the utility room, a detail shot of the panel label, a context shot showing tight clearance, and narrate the required upgrade with labor notes. This structured data enables precise automated proposal generation.
Your Implementation Roadmap
- Adopt the Four-Shot Discipline: Rigorously apply the photo sequence on your next three site visits. Consistency trains both your habit and the AI.
- Script Your Audio Checklist: Keep the key information points (Identification, State, Action, Labor) on a quick-reference card or phone note. Use them verbatim.
- Leverage a Unified Tool: Utilize a digital notebook tool like Obsidian to store and link your media. Its purpose is to keep photos and voice notes organized in one searchable project folder, creating a clean data package ready for AI processing.
By transforming your site capture into a structured process of visual context and narrative detail, you turn raw observations into AI-ready intelligence. This discipline is the foundation for automating accurate material lists, labor estimates, and professional proposals, directly from the field.
![Defluffer - reduce token usage 📉 by 45% using this one simple trick! [Earthday challenge]](https://media2.dev.to/dynamic/image/width=1000,height=420,fit=cover,gravity=auto,format=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fiekbgepcutl4jse0sfs0.png)
