FieldDraft — overnight survey draftsman for small US land-survey firms

1. One-liner

Turns a day’s raw field data into a standards-compliant draft plat and legal description by morning.

2. Trend signal — why now?

Three things converged in the last 12 months:

1. The drafting bottleneck is now an industry, not a complaint. A whole tier of offshore shops (Aidedo, Gsource, TIVOP, Survey Drafters LLC) sells nothing but “clear your 4-week drafting backlog” — TIVOP literally promises a recordable drawing back in 24–48 hours and opens with “if your queue is over 2 weeks right now, that is exactly the conversation I want to have.” When the workaround has its own SaaS-style landing pages and flat per-sheet pricing (~$185/D-size sheet), the pain is validated by people already spending money on it.
2. The labor supply is structurally broken. Average age of a licensed US land surveyor is ~58; only 14% are under 34; 2–3 retire for every new entrant (NSPS / BLS). Firms can’t hire the CAD techs who’d clear the backlog, so they either outsource or fall behind. This isn’t cyclical — it gets worse every year.
3. The AI tooling crossed the line from “review” to “produce.” 2025–26 LLM+geometry tools (Bunting Labs reads legal descriptions → DXF; Survey XYZ critiques plats; Legal-Aid generates descriptions from drawn geometry) prove each sub-step works. Nobody yet does the actual painful labor: raw field codes/point file → drawn linework + annotated plat + prose legal description in the firm’s own CAD standards.

Provenance:

• Signal 1: Surveyor burnout / office work piling up after field days; CAD “made for civil engineers, survey drafters an afterthought” — https://smartlandsurvey1.wordpress.com/2025/09/09/burned-out-construction-surveyors-reddit-reveals-crisis/ — 2025-09-09
• Signal 2: US land surveyor shortage, avg age 58, 14% under 34, 2–3 retiring per entrant — https://360talentavenue.com/why-are-land-surveyors-suddenly-becoming-scarce-to-hire/ and https://www.bls.gov/ooh/architecture-and-engineering/surveyors.htm — 2025
• Signal 3: Outsourced drafting industry priced ~$185/sheet, 48hr SLA, targeting “4-week backlogs” — https://www.tivop.com/ and https://aidedo.com/services/boundary-survey-drafting-service.html — 2026 Category: Workflow automation

3. The opportunity

The incumbent isn’t a software company — it’s a fleet of offshore CAD drafting shops. They’re slow (24–72 hr round trips), they need detailed written instructions and back-and-forth markups, they don’t learn your firm’s CAD standards (you re-correct the same things every job), and the unit cost is fixed labor (~$185/sheet) that never drops.

An AI-first product collapses that loop. Upload the LandXML / point file / field-code list and a photo or scan of the field book; get back a draft DWG with linework, symbols, and a metes-and-bounds legal description rendered in your template and CAD standards, before the next field day — for a price that undercuts a per-sheet shop while being 20× faster. The surveyor still reviews, corrects, signs and seals (legal requirement, and the moat). We’re replacing the drafting labor, not the licensed judgment.

10× better, specifically: turnaround goes from days to overnight; the tool memorizes the firm’s block library, layer standards, and description boilerplate so corrections compound down instead of repeating; and marginal cost per sheet trends toward inference cost, not offshore wages.

4. Target market

• Primary customer: Owner / licensed PLS at a 1–20 person US land-survey firm doing boundary, topo, ALTA, and subdivision-plat work — the exact firm TIVOP targets (“1–20 employees, queue over 2 weeks”).
• Why they buy: “I’m a month behind on drafting, I can’t hire a tech, and the offshore shop still needs me to babysit every sheet.” Deliverable turnaround is the thing clients scream about and the thing that caps the firm’s revenue.
• Rough TAM reasoning: ~10,000+ small/solo survey firms in the US (NSPS, state board rolls). Even 3,000 reachable firms × $6K/yr ACV ≈ $18M serviceable. Sub-$5M ARR needs a sliver.
• Why now for them: The retirement cliff is biting this year — the people who used to draft are gone and not coming back. Outsourcing got normalized during 2023–25, so paying an outside party to draft is no longer culturally novel; switching from an offshore shop to an AI tool is a price/speed decision, not a behavior change.

5. Product sketch (MVP)

• Upload field data: LandXML, raw CSV point file with field codes, or a photo/scan of the field book.
• Auto-generate draft linework + survey symbols from field codes (linework connectivity, curves, lot lines).
• Generate a metes-and-bounds legal description in prose, with closure check and error-of-closure report.
• Render output in the firm’s uploaded CAD template (title block, layer standards, block library).
• “Standards memory”: every correction the surveyor makes teaches the firm-specific profile for next job.
• Output a review-ready DWG + PDF + the legal description text — explicitly watermarked “DRAFT — for licensed review.”
• Closure / geometry sanity flags surfaced up front so the PLS reviews the risky parts first.

6. AI angle — what’s load-bearing

Remove the AI and there is no product — it’s just a CAD template. The load-bearing work is: (a) interpreting messy, inconsistent field codes and field-book shorthand into correct linework intent (every crew codes slightly differently — this is judgment, not lookup); (b) generating a legally-phrased metes-and-bounds description from geometry that reads like a human surveyor wrote it; (c) learning a specific firm’s drafting standards from corrections so output converges. That middle layer — fuzzy field shorthand → correct draftsman intent → firm-standard output — is exactly what an LLM+geometry pipeline now does and what no current tool (review-only or geometry→text-only) attempts end to end.

7. Localization angle (if any)

N/A — this is a US-first play. The wedge is US survey conventions (PLSS, state plat statutes, ALTA/NSPS standards) and the US labor-shortage economics. International expansion (AU/UK cadastral) is a later, separate product with different rules; forcing a localization angle now would dilute the wedge.

8. Business model — path to $1M–$5M ARR

• Pricing: Tiered SaaS with a usage cap. $399/mo solo (up to ~20 deliverables), $799/mo firm (up to 60), $1,499/mo for high-volume — plus per-sheet overage. Anchored below per-sheet offshore ($185/sheet) on a per-deliverable basis while being instant.
• ACV: ~$6,000/yr blended.
• Rough math to $1M ARR: 170 firms × $799/mo × 12 ≈ $1.6M. Or 280 firms at the $399 tier ≈ $1.3M.
• Rough math to $5M ARR: ~700 firms at blended $6K ACV, or ~450 firms with overage/upsell pushing ACV to ~$11K. Both well inside the 10K-firm TAM.
• Expansion path: overage on volume; add ALTA-specific module, construction-staking sheets, deed-research/title-search assist; per-seat for multi-drafter firms. ACV grows by deliverable type, not by squeezing more seats.

9. Go-to-market wedge — first 100 customers

• Steal the outsourcers’ customers. Offshore drafting shops publicly list/showcase client work and target the same 1–20-person firms. Build a list of firms visibly outsourcing (case studies, LinkedIn, “we use [shop]” mentions) and pitch: same deliverable, overnight, cheaper, no markup babysitting.
• State board / NSPS rolls. State licensing boards publish licensed-PLS rosters with firm names. Scrape ~3,000 small firms, send a personalized before/after: their own sample field file → a generated draft plat in 24h. Demo is the sales pitch.
• RPLS.com, Land Surveyors United, r/Surveying. The drafting-backlog complaint threads are active and specific. Show up with a real generated deliverable, not a teaser. Convert 2–5% of the engaged.
• State surveyor association conferences (every state has one; cheap booths). The audience is exactly the buyer and they all feel this pain — one good live demo of “field file in, draft plat out” travels by word of mouth in a tight profession.
• Target: 30 paying firms in 90 days via direct outreach #1–2; 100 within 7–9 months adding community + conferences.

10. Build complexity — justification

Medium. Off-the-shelf: LLM APIs, geometry libraries, DWG/DXF read-write libraries, standard web stack. Custom work that takes real engineering: a reliable field-code → linework engine (curves, connectivity, control), a closure-correct legal-description generator, and the firm-standards learning loop. Plus genuine domain expertise — you must understand survey field workflow or the output is plausible garbage that no PLS will trust. Realistic v1 for a 2–3 person team (one with survey domain depth): ~4–5 months.

11. Gating checklist

Gate	Pass?	Note
Legal in target market	✅	Produces a draft; licensed PLS reviews, signs, seals. We never determine boundaries or certify.
Ethical — no harm / dark patterns	✅	Hard “DRAFT — for licensed review” watermark; closure errors surfaced, not hidden. Augments, doesn’t impersonate, the PLS.
Market exists (evidence above)	✅	Entire offshore drafting industry + structural labor shortage + active complaint threads.
1–5 person team can build this	✅	2–3 people, one with domain expertise, ~4–5 months.
Launchable with <$50K / ₹40L	✅	Inference + dev time. No capex, no licenses to acquire.

All five pass.

12. Feasibility score

Axis	Weight	Score	Notes
Problem intensity	20	16/20	Firms lose revenue to turnaround caps and already pay outsourcers monthly. Real, repeated, money-on-table — short of “hair on fire daily” only because they have a working (slow) workaround.
Demand evidence	15	12/15	Strong: a priced outsourcing industry, structural labor data, active complaint threads. Docked — most evidence is the workaround market, fewer clean verbatim “I’d pay for this exact product” quotes.
Build feasibility	15	10/15	Doable in ~4–5 months but the field-code→linework + closure-correct description engine is non-trivial and the trust bar is high. Not a 6-week solo build.
Distribution clarity	15	11/15	Named lists (board rolls, outsourcer clients, RPLS/LSU), demo-as-pitch, tight word-of-mouth profession. Conversion unproven; not a 2-week sprint.
Revenue mechanics	15	12/15	Pricing benchmarked directly against $185/sheet incumbents; ACV and customer counts conservative within TAM. One assumption (PLS trusts AI draft enough to pay monthly) still a guess.
Time to first revenue	10	7/10	4–8 weeks post-launch realistic via direct outreach + instant demo deliverable. Not pre-sold day one.
Defensibility	10	8/10	Firm-standards learning loop = compounding workflow lock-in; survey-domain correctness is a real barrier to casual clones. Not patent-grade but sticky.
Total	100	76/100

13. Qualitative modifiers

Founder-fit tags

technical-heavy · domain-expertise-required — needs solid AI/geometry engineering AND someone who has actually drafted survey deliverables or worked at a survey firm. Without the domain half, output won’t clear the PLS trust bar and the company dies on credibility.

Key assumptions to validate (3–5)

1. Assumption: A licensed PLS will trust and pay monthly for an AI-generated draft they still must review. How to test: Run 15 firms’ real field files through a hand-assisted prototype; measure how much correction time they spend vs. their offshore round-trip, and whether they say “I’d pay for this.”
2. Assumption: Field-code → linework can be made reliable enough across firms’ inconsistent coding to be net-time-saving, not net-rework. How to test: Collect field files from 10 firms; measure draft accuracy and surveyor correction minutes per sheet.
3. Assumption: Per-deliverable price below offshore ($185/sheet) still yields healthy margin after inference + review-tooling cost. How to test: Cost a full deliverable end-to-end at target model usage; confirm gross margin >70% at $399–799 tiers.
4. Assumption: Small firms will switch from an established offshore shop. How to test: Cold-pitch 50 firms known to outsource; measure reply + trial rate.

Risk flags

1. Trust / liability risk: Surveyors are conservative and personally liable under seal. One bad public mistake poisons the brand in a small, gossipy profession. Mitigate with aggressive closure flagging and “draft only” framing — but this gates adoption speed.
2. Platform dependency: Heavy reliance on third-party LLM/geometry APIs for the core; pricing or capability shifts hit margin and quality directly.
3. Incumbent response: Offshore shops could bolt AI on and keep their existing client relationships and price wars. Our edge is speed + standards-memory, which they’re structurally slow to match — but it’s a race, not a moat-by-default.
4. Market timing: Survey-domain AI tools (Survey XYZ, Bunting Labs, Legal-Aid) are landing now — the category is forming. Good (validates demand) and bad (window won’t stay open; move within ~12 months).

14. Structured verdict

Score:                  76/100
Verdict:                GO
Confidence:             Medium
Best-fit builder:       Technical founder + survey-domain co-founder/advisor (ex-survey-firm drafter or PLS)
Time to revenue:        6–10 weeks post-launch (4–5 months to launch)
Capital to launch:      $15–35K (inference + dev time; no capex)
Top 3 assumptions to validate first:
  1. PLS will pay monthly for an AI draft they still review — 15-firm prototype trial, measure correction time vs offshore
  2. Field-code → linework is net-time-saving across inconsistent firm coding — 10-firm accuracy/rework test
  3. Per-deliverable price under $185/sheet still yields >70% gross margin — full end-to-end cost model
Kill criteria:
  - Abandon if <20% of 50 outsourcing-firms cold-pitched will run a free trial deliverable
  - Abandon if surveyor correction time on AI drafts exceeds 50% of doing it from scratch in the 15-firm trial
  - Abandon if gross margin at the $399/$799 tiers falls below 55% after real inference costs

15. Next step — 1-week validation sprint

• Day 1–2: Pull 50 small US survey firms that visibly outsource drafting (offshore-shop case studies, LinkedIn, board rolls). Collect 8–10 real anonymized field files from friendly firms / forum contacts.
• Day 3–4: Hand-build (Wizard-of-Oz, human-in-loop) draft plats + legal descriptions from 5 of those files in the firm’s template. Send each firm its own before/after.
• Day 5: Cold-pitch the 50. Decide go/no-go on a falsifiable bar: ≥10 firms (20%) request a paid trial AND ≥3 of the 5 hand-built deliverables come back rated “I’d review-and-use this” with correction time under half of from-scratch. Below that bar, the trust gap is too wide — revisit or kill.