WorkSight

Logo
Tile

Inspiration

For any organization and their sales force, finding the ideal contacts to engage with is one of the most challenging and time consuming tasks that offers little reward on the front end. Building effective prospecting lists can be a cumbersome process and requires sales reps in any industry to build lists and then verify them. For outdoor contractors this often requires using Google Maps to find the location of a property and then manually copy and paste information such as address and phone number. Then they have the difficulty of determining if the lead is sound, and finding a decision maker. Contractors can also use scraping tools, and then paid for contact databases that usually have outdated information. We were inspired to build a tool that combines all of these functionalities into one place and delivers a comprehensive solution that not only finds quality leads but accelerates speed to lead, and engages with those potential customers to shortens the gap from discovery to close.

What it does

When a client signs up for WorkSight they fill out a client profile form that our AI analyzes to determine the sweet spot for their business and its capabilities based on their size of operation and ideal customer profile. The client then searches for properties using an integrated map tool which finds properties that meet the search parameters. The WorkSight AI then gains an InSight on any properties that the client is interested in to enrich the contact data for the property. Once the parcel's contact information has been enriched the client can then use our integrated AI agent to do a further deep dive into the properties pertinent information and save the results to the contact. The AI learns over time what is important to the client to help speed up the chain of thought reasoning and deliver more pertinent information faster.

How we built it

We are building the app with Cursor, Gemini grounded search, Google Maps, VM, and Claude code.

Challenges we ran into

There are many challenges with integrating parcel identification, records of ownership, business profiles, AI analysis with segmentation models, virtual machines, and outreach tools.

Accomplishments that we're proud of

We built a production AI platform that discovers, scores, enriches, and markets to commercial properties for trade contractors — using zero manual prospecting. Our multi-model computer vision pipeline (SAM 3, DINOv3, YOLOv9, GPT-Image-2) segments pavement from satellite imagery at pixel-level precision, enabling automated surface takeoffs in seconds that previously took hours of manual site visits. We integrated the Gemini API with Google Search grounding to autonomously research and identify decision-maker contacts for every property, complete with cited sources and confidence scores. Our Postcard Studio generates AI-designed direct mail with unique property-specific QR codes that link to personalized intake portals — turning every mailed piece into a trackable, closed-loop engagement. We have live CRM integrations (Simpro, OneCrew), active enterprise level clients, active community partnerships with industry podcasts and trade groups, and a fully operational pipeline that transforms a contractor's lead generation from a tedious guessing game into an automated intelligence system.

What we learned

We learned that the trades industry is far more underserved by AI than we initially realized — contractors have been paying for generic lead lists and Google Ads for years, but no one has built a system that starts with satellite analysis and works forward from there. We also discovered that contractors need full control over the AI's decisions, not a black box — giving them scoring controls, campaign rules, and natural language refinement over AI-generated postcards was what turned curiosity into trust. Most importantly, we learned that the real moat isn't the AI models themselves, but the recursive learning loop: the more a contractor uses the platform, the better WorkSight becomes at finding their specific kind of lead, making the system more valuable every single day.

What's next for WorkSight

We believe in a world of abundance where human ingenuity and talent are not wasted on mundane tasks like prospecting. Every commercial property will have its own AI running the property and when maintenance needs come up, those agents will handshake with WorkSight to land workorders without human intervention. We are expanding beyond paving and asphalt into adjacent verticals — roofing, solar, landscaping, and concrete — where the same satellite-visible surface condition data unlocks identical value for contractors. We are deepening our CRM integrations with Jobber and GoHighLevel so WorkSight becomes the lead generation engine inside the tools contractors already use every day. On the product side, we are moving toward fully autonomous agentic outreach — where WorkSight doesn't just find leads, but manages the entire campaign lifecycle, adjusting messaging, timing, and channels based on real-time engagement signals like QR scans and email opens. We are investing in our fine-tuned segmentation models using real contractor job data (KMZ polygons from actual completed projects) to make our surface takeoffs even more accurate — turning every job a contractor wins into training data that makes the system smarter.