Aiolum Browse stories
All stories
AI plan review for development approvals: checks site and building plans against local codes before and during city review · United States

AutoSitu: Product, team and how it was built

Small teamFundedB2BConstruction

AutoSitu (AI plan review for development approvals: checks site and building plans against local codes before and during city review, United States): 25K+ developments analyzed (company figure), November 2026. Evidence grade E: no revenue figure was checked. Team: 2 co-founders; team size 2 on the Y Combinator page, November 2026. Launched 26 January 2026; still running.Added 6 November 2026

Outcome
25K+ developments analyzed (company figure), November 2026
Evidence
E no revenue figure was checked
Launch
26 January 2026
Status
Still running
Team
2 co-founders; team size 2 on the Y Combinator page, November 2026
Outside funding
Funded: Y Combinator (Winter 2026), a seed round led by Pi Labs, and Robinhood Ventures Fund II (August 2026); amounts not found in the sources read
Main channel
Not stated
No revenue figure

What is not verified

No revenue figure was found. No payment-processor profile, funding amount, filing, press report or founder interview with revenue was found, and a web search on 6 November 2026 found no city council record or contract award naming AutoSitu. The headline counter (more than 25,000 developments analyzed) and the other home-page counters are the company's own figures, which its For Cities page calls "verified platform totals" without saying who verified them. These are large numbers for a two-person company ten months after its launch post; a reader could look for a named city pilot, a procurement record, or a demo day announcement with figures. This story is grade E: the business is real and its product, price and customers are described from public pages, but no revenue figure was checked. Aiolum shows it as an idea worth studying, not as proof that it makes money.

AutoSitu sells plan review software. Plan review is the step in which a city's planners and engineers check a project's drawings against the local zoning code (the rules for use, height, setbacks and density on each parcel) and the building and fire codes before they approve it. AutoSitu reads the plan set, runs the repeatable checks, and cites the code section behind each finding; a human reviewer makes every decision.

The founders are Xuanshu (Asher) Lin, who studied urban planning and architecture at the University of Michigan, and George Zhai, who worked on autonomous cars, boats and robots. The company was founded in 2025, joined Y Combinator's Winter 2026 batch, and posted its launch on 26 January 2026. A seed round led by Pi Labs followed, and in August 2026 the company joined the first portfolio of Robinhood Ventures Fund II.

No revenue figure was found. The headline is the company's own counter on its home page: more than 25,000 developments analyzed. This is grade E: the business is real, live and sold by contract, but its own money is not public.

AutoSitu: 25K+ developments analyzed (company figure), November 2026. AI plan review for development approvals: checks site and building plans against local codes before and during city review, United States, team of 2 to 10, evidence grade E.
The story in one card. Share it with the address of this page.
01

How much revenue does AutoSitu make?

Each row shows the evidence grade and the source numbers from the list at the end of this page. The grades are explained in How Aiolum verifies.

A two-founder plan-review product for cities and developers, live and sold by order form after a Y Combinator launch, with no published revenue
FactValueGradeHow we knowSources
Status, November 2026live; sold by order form after a demoCHome page and Terms of Use (last updated 24 October 2026), read 6 November 2026. The site has a Sign in page for customers.[3, 7]
Developments analyzed (company counter)more than 25,000CHome page counter, read 6 November 2026. The For Cities page calls such counters "Verified platform totals across the submittals AutoSitu has reviewed — not projected ROI". No outside party has checked them.[3, 6]
Other company counters8.7 million+ requirements checked; 420,000+ issues identifiedCHome page. The same page states US$18 billion+ of "development value evaluated", which is the value of customers' projects, not AutoSitu's revenue.[3]
Founded2025CY Combinator company page.[1]
Team, November 20262 people, both co-foundersCY Combinator company page lists team size 2, with George Zhai and Xuanshu (Asher) Lin.[1, 4]
Launch post26 January 2026CLaunch YC post by the founders, created 26 January 2026.[2]
FundingY Combinator (W26); seed round led by Pi Labs with Y Combinator, Detroit Venture Partners and Michigan Rise; Robinhood Ventures Fund IICCompany blog post of 6 August 2026. No amounts are stated in the sources read.[5]
Results claimed for users50% to 70% fewer city comments; resubmittals cut from 2 to 3 down to 0 to 1CFounders' launch post. These are customers' outcomes as the founders state them, not AutoSitu's revenue, and no customer is named.[2]
Who buys itcity planning departments, developers, architects and engineers, contractors, planning firmsCHome page menus. An independent research database (Chain Drain) writes that it primarily sells to government agencies.[3, 9]
Pricesnot publishedCNo pricing page; demo only.[3, 7]
Time from the launch post to November 202610 monthscalcCalculated: January 2026 to November 2026.[2, 3]

How to read this table: every value is a statement by the company on its own pages or on its Y Combinator pages, grade C, or a third-party research note that read the same pages. A "calc" value was computed on this page. No value is grade A or B, and no revenue figure exists.

AutoSitu prices in November 2026: not published (the site offers a private demo)
PlanPer month, billed yearly (US$)Per month, billed monthly (US$)Limit
Contract after a demonot publishednot publishedorder form or subscription agreement; pilots for city departments

The site has no pricing page (/pricing returns 404, checked 6 November 2026), and every call to action is "Schedule Private Demo" or "Book Demo". The Terms of Use say fees are set by "the applicable order form, invoice, subscription agreement, statement of work, or other written agreement". The For Cities page says department benchmarks come from "your own pilot".

02

When did AutoSitu launch, and how did it grow?

From a Y Combinator launch post to a seed round and a Robinhood fund listing, in about ten months of public pages. Each entry carries its grade.

  1. Before 2025CLin works on more than 20 public and private planning and architecture projects in the United States; Zhai builds automation systems for autonomous vehicles and robots. [1, 2]
  2. 2025CThe two founders start AutoSitu. [1]
  3. 26 January 2026CLaunch post on Y Combinator: plan review that "surfaces what a city is likely to flag before you submit". [2]
  4. 28 July 2026CThe Chain Drain research database adds a dossier on AutoSitu and classes it as plan-review software sold mainly to government agencies. [9]
  5. 6 August 2026CBlog post: AutoSitu joins the first portfolio of Robinhood Ventures Fund II, after a seed round led by Pi Labs. [5]
  6. November 2026CThe site describes an "AI operating system for the built world" with six modules and a home-page counter of more than 25,000 developments analyzed. [3]
03

Why did AutoSitu work?

Four causes, in the order they mattered. Each one is a reusable lesson.

01

Domain knowledge

One founder worked in planning and architecture on both sides of the review table, the other in automation. The launch post says the product started from one observation: the issues that cities flag are predictable, but each one is checked by hand.

02

A human makes every decision

The software checks and cites; the city reviewer decides and signs. The For Cities page says "Staff remain the decision-maker", which is what a public agency needs before it can use such a tool.

03

Two sides of one review

The same checks are sold to the applicant (to file cleaner plans) and to the city (to review faster). Each side's use makes the other side's review shorter.

How it was built

The company has not published its technical stack in detail. These details come from its own pages and are grade C, except the page framework, which is visible in the page source. [6, 8, 3]

  • Plans, drawings and PDFs are processed by "AI service providers" and "AI model" providers, as the privacy policy says; it names no provider.
  • Each finding is linked to the adopted ordinance or code section and the sheet it was found on.
  • Resubmittals are compared with the earlier set, so reviewers see what changed and which comments were answered.
  • The website is a Next.js application (its page source loads files from /_next/static/), and demos are booked through a Google Calendar booking page.

Tools named in the sources

Each one opens its page in the Tools tab, with every story that uses it.

04

Who tried this before AutoSitu, and why did they fail?

Three companies in our database of failed startups promised to make building faster or design easier with technology, and failed. The values below are from that database and are not independently verified.

Died 2015 to 2022Burned US$2.6B

Katerra

The promise

Vertically integrate construction, from design to factories to the building site.

Why it died

It built factories before it had customers and bought companies until it went bankrupt.

Unit Economics
Died 2017 to 2020Burned US$17M

Social Construct

The promise

Build housing quickly and cheaply with a technology-led construction process.

Why it died

It underestimated the regulatory work and the cost of scaling, and ran out of money during the pandemic.

Ran Out of Cash
Died 2011 to 2025Burned US$550M

Kujiale

The promise

Cloud software for 3D interior design and furniture planning.

Why it died

Consumers did not need a design tool, and newer AI tools made its technology old.

No Market Need

The failed startups come from our database of 1,749 dead startups, retold here in our own words. Its entries are AI-assisted summaries of public sources, so these cards carry no evidence grade.

What changed

Katerra and Social Construct tried to change how buildings are built, which needed factories, crews and large amounts of capital. AutoSitu changes only the paperwork before construction: the check of drawings against written rules, which is software work.

Language models can now read long code documents and plan sheets and turn them into checks with citations. That lets a team of two sell a first-pass reviewer to cities and design firms, a job that earlier needed custom rule engines written for each city.

05

Can one person copy AutoSitu now?

Not alone. The method, yes.

Selling to city governments needs pilots, procurement and trust, and the product must model each city's code; that is heavy for one person. The method is open: find a document check that experts repeat by hand on every case, and sell software that does the first pass with citations.

Where it is still open: Other rule-based reviews with public codes: fire inspections, accessibility audits, environmental permits, utility interconnection applications.

  1. 1Work in the review process first, on both sides if possible.
  2. 2Pick checks that are written in public rules, so each finding can cite its source.
  3. 3Keep a person as the decision-maker, so public buyers can adopt it.
  4. 4Sell the same check to the applicant and to the reviewer.

What could go wrong for AutoSitu?

The risks are the mirror image of the causes above.

What is not verified

No revenue figure was found. No payment-processor profile, funding amount, filing, press report or founder interview with revenue was found, and a web search on 6 November 2026 found no city council record or contract award naming AutoSitu. The headline counter (more than 25,000 developments analyzed) and the other home-page counters are the company's own figures, which its For Cities page calls "verified platform totals" without saying who verified them. These are large numbers for a two-person company ten months after its launch post; a reader could look for a named city pilot, a procurement record, or a demo day announcement with figures.

Logo wall and testimonial

The home page shows "Join peer cities" followed by the City of San José, Seattle, Long Beach, Boston, Los Angeles, Dallas, San Francisco and others, and the heading "Trusted by visionary municipalities and firms". No page read says which of these cities are customers, users of a pilot, or only invited peers. The one quotation on the home page is from an unnamed "VP of Design & Construction, National real estate developer".

Competition and dependence

Like Avoice (drawing checks for architecture practices) and Blueprint Crusher (estimating from tender documents) on this site, AutoSitu reads construction documents with AI; it differs by selling to city review departments. The Chain Drain dossier lists CivCheck, Symbium and Archistar as competitors in AI plan review. The privacy policy says project materials may be processed by "AI model" providers but names none, so the dependence on a model provider cannot be assessed.

Public-sector stakes

Findings feed permit decisions. The company says a human reviewer decides every finding, and its Terms of Use say third-party data such as parcel records and code references "may be incomplete, delayed, inaccurate, or unavailable".

Sources

The grade next to each source is the best grade that the source can support. "Read in full" means that the text of the page was downloaded and the numbers were checked in it.

  1. C Y Combinator company page: AutoSitu
    Read in full (page data downloaded 6 November 2026). Source for founding year, team size, founders, their backgrounds and the batch.
  2. C Launch YC, 26 January 2026: AutoSitu, AI-native Development Plan Review for Faster Approvals
    Read in full. Founders' launch post. Source for the problem, the claimed results for users and the founders' backstory.
  3. C AutoSitu home page, read 6 November 2026
    Read in full. Source for the modules, the counters, the city logo wall, the testimonial and the demo-only sales.
  4. C About AutoSitu
    Read in full. Source for the mission and the founding team.
  5. C AutoSitu blog, 6 August 2026: AutoSitu Joins the Robinhood Ventures Fund II Portfolio
    Read in full. Source for the seed round investors and the Robinhood fund.
  6. C AutoSitu for Cities
    Read in full. Source for how city reviewers use it, the human decision step and the pilot wording.
  7. C AutoSitu Terms of Use, last updated 24 October 2026
    Read the sections on fees, third-party services and governing law. Source for the order-form sales and the Michigan governing law.
  8. C AutoSitu Privacy Policy, last updated 24 October 2026
    Read in full. Source for the use of AI model providers and the United States base.
  9. C Chain Drain research database (Bryan Boyer): AutoSitu dossier, added 28 July 2026
    Read in full. A third-party research note built from the company's pages; source for the buyer classification and the named competitors.
  10. C Database of failed startups, local export
    Local export. AI-written summaries, not independently verified. Source for the failed predecessors.

Run this company? Send corrections or a verified figure to aiolum@livyden.com

Back to all stories