Six sectors we have actually shipped into.
Domain knowledge is not a slide — it is knowing that a retailer will never accept a refused sale, that an estimator needs to know which lines to check, and that a volunteer has to be able to update the site in two years.
Retail & commerce
Physical stores that also sell online, and refuse to run two systems to do it.
Retail software fails at the seams: stock that disagrees between the shelf and the website, pricing that has to be maintained in four places, a register that stops working when the connection does. We build retail systems as one catalog with many surfaces.
What we already know here
- Offline-capable point of sale and hardware integration
- Per-location stock, purchase orders and vendor receipts
- Marketplace publishing with stock and price kept honest
- Loyalty, store credit and the accounting behind both
- Tiered and wholesale pricing with server-side precedence
Construction & AEC
Contractors, subcontractors and specialty trades who bid, build and bill.
The AEC stack is unusually fragmented and unusually document-heavy, which makes it one of the few industries where AI changes the economics rather than the marketing. Drawings are data; most software treats them as attachments.
What we already know here
- Quantity takeoff from drawings, with confidence scoring
- Estimating that converts into projects without re-keying
- AIA progress billing, retainage and lien waivers
- Field capture that tolerates bad connectivity
- Document control across revisions people actually build from
Consumer products
Apps where the second user is a guest who will never make an account.
Consumer software lives or dies on the first thirty seconds and on what you do not require. Every field, every signup wall and every permission prompt is a place people leave. The engineering problem is making a rich product feel like no commitment.
What we already know here
- No-account flows that still capture structured data
- Multi-language from the first release, not retrofitted
- Media storage with retention rules the product honours
- Share surfaces: link, QR, email, SMS, messaging apps
- Payments for people who will use the product once
Professional services
Consultancies and technology firms selling expertise into enterprise buyers.
When the product is judgment, the website is doing the qualification. The work is making a broad capability set legible to a specific buyer without collapsing into a list of everything.
What we already know here
- Capability architecture that survives a broad service mix
- Industry positioning without brand fragmentation
- Credibility signals aimed at procurement, not consumers
- Content structures that hold up as the practice grows
Local & trade services
Businesses whose customer is searching on a phone with an urgent problem.
Nobody browses for an appliance repair company. They search once, in a hurry, and call whoever answers the question fastest. The site has one job: be the obvious call. Everything decorative competes with that.
What we already know here
- Service-area structure built around real search behaviour
- Trade credibility surfaced above the fold
- Enquiry and booking flows built for urgency
- Multi-city coverage without thin duplicate pages
Community & nonprofit
Organizations whose website is a schedule, a welcome and a place to belong.
Community organizations are usually run by volunteers with no time and no budget for maintenance. The measure of success is whether someone can still update it in two years without calling anyone.
What we already know here
- Recurring schedules that stay correct without weekly edits
- Content that volunteers can maintain unaided
- Bilingual and heritage-language communities
- Low-cost hosting with effectively no ongoing burden
Not on this list?
Sector experience is useful, not decisive. The engineering that makes a takeoff tool work is the engineering that makes anything document-heavy work.