AArchPermit
How ArchPermit compares

Most tools do one piece. ArchPermit does the whole path.

ArchPermit is not an autonomous AI architect, not a pure code-lookup tool, and not a permit-ops workflow alone. The wedge is simple: no existing product goes from a messy brief to an architect-reviewed, jurisdiction-specific, citation-grounded, permit-ready package. ArchPermit does.

Where the other categories stop

Not an autonomous AI architect

ArchPermit automates the production work, but a licensed architect keeps responsible control and the seal wherever the law requires it. Autonomy stops at the things that carry legal accountability.

Not a pure code-lookup tool

Code lookups tell you the rules. They don’t resolve your jurisdiction from an address, read your documents, assemble a draft package, or route it through a licensed review.

Not a permit-ops workflow alone

Expediter and permit-ops tooling coordinates filings and people. ArchPermit does that on top of the grounded, cited intelligence that turns a brief into the package in the first place.

Feature by feature

An honest, side-by-side look at what each category actually does for a real submittal. Where another approach can do part of the job, we say so.

Jurisdiction resolution from an address

Work out the Authority Having Jurisdiction (plus flood and elevation context) from a plain address.

ArchPermitYes
Automatic from the address — resolves the AHJ before building the checklist.
DIY code research / code-lookup toolsNo
You pick the code book yourself; no address-to-jurisdiction step.
Permit expediters / permit-ops workflowsPartial
A person knows the jurisdiction, but only once you’ve engaged and paid them.
Generic AI tools that “guess”No
Will name a jurisdiction confidently, with no guarantee it’s the right one.

Requirements grounded in cited gov sources + last-reviewed date

Every requirement carries a citation to a government source and the date it was last reviewed.

ArchPermitYes
Each item links a government source and shows its last-reviewed date.
DIY code research / code-lookup toolsPartial
Points you at the codes, but you assemble and date-check the checklist yourself.
Permit expediters / permit-ops workflowsNo
Knowledge lives in the expediter’s head; no cited, self-serve checklist you keep.
Generic AI tools that “guess”No
Answers read authoritative but are usually uncited and may be stale or invented.

Says so when unsure (fail-closed, no fabrication)

When a source conflicts, an address can’t be placed, or confidence is low, it stops and says so.

ArchPermitYes
Fail-closed by design — flags uncertainty for a human instead of inventing a requirement.
DIY code research / code-lookup toolsPartial
It’s just reference material — it can’t tell you what it doesn’t cover.
Permit expediters / permit-ops workflowsPartial
A good expediter will say “I’m not sure” — but it depends on the individual.
Generic AI tools that “guess”No
Hallucination-prone — fills gaps with confident, fabricated answers.

Reads your documents & extracts permit-relevant facts

Reads a survey, title report, or soils report and pulls the parcel, zoning, and foundation facts.

ArchPermitYes
Document intelligence extracts the facts and links them to the right checklist items.
DIY code research / code-lookup toolsNo
Static reference; doesn’t ingest or read your project documents.
Permit expediters / permit-ops workflowsPartial
A human can read your documents, but it’s manual, billed, and not self-serve.
Generic AI tools that “guess”Partial
May summarize a pasted file, but extraction is unverified and not tied to a checklist.

Generates a draft package

Assembles a jurisdiction-specific draft permit package you can iterate on.

ArchPermitYes
Produces a Draft permit-package PDF, assembled against the cited checklist.
DIY code research / code-lookup toolsNo
Tells you what’s required; you produce the package separately.
Permit expediters / permit-ops workflowsPartial
Coordinates and files a package, but the production is a slow, expensive human service.
Generic AI tools that “guess”No
Can draft prose, but not an assembled, jurisdiction-specific submittal set.

Licensed-architect review + seal gate (enforced in code)

Nothing is permit-ready until an independent licensed architect reviews and seals it.

ArchPermitYes
The seal gate is enforced in code — you can never seal your own project.
DIY code research / code-lookup toolsNo
No review and no seal — it’s reference, not a responsible professional.
Permit expediters / permit-ops workflowsPartial
Files with the city, but the licensed design seal is a separate engagement you arrange.
Generic AI tools that “guess”No
No licensed gate — nothing stops unreviewed output from looking final.

Audit trail

Every resolution, status change, document read, and review decision is recorded.

ArchPermitYes
Full audit trail — you always know what ArchPermit did and why.
DIY code research / code-lookup toolsNo
Nothing is recorded; it’s a lookup, not a system of record.
Permit expediters / permit-ops workflowsPartial
Email threads and filing receipts — no structured, queryable history.
Generic AI tools that “guess”No
Chat logs at best — no per-decision evidence trail.

Yes = does this directly · Partial = can do part of it, or only via a human/manual step · No= not what the category does. Categories are described generically; we don't compare named products or invent their pricing.

The difference is the licensed-review gate.

Speed is easy to copy. Accountability is not. Everything ArchPermit produces is marked Draft — pending licensed review until an independent licensed architect reviews and seals it. You can never seal your own project — the rule is enforced in code, not just promised in copy. That is the line between a fast guess and something you can put in front of a building department.

Draft until sealed

Nothing is presented as permit-ready. Every artifact stays labelled Draft — pending licensed review until a licensed professional signs off.

Independence enforced in code

You cannot review or seal your own project. The separation between author and reviewer is a code-level rule, not an honor system.

Grounded and auditable

Cited sources with last-reviewed dates, plus a full audit trail of every resolution, document read, and review decision — so the package can withstand scrutiny.

Comparisons describe categories of tools and services generically, based on how each approach typically works; they are not claims about any specific named product. We don't quote competitor pricing — it varies by provider, scope, and jurisdiction. Any product or company names that may apply are the trademarks of their respective owners, and their use does not imply affiliation or endorsement.