Company profile · 3 min read
Alkali: AI steel takeoff, nesting and revision control
Alkali helps structural-steel teams estimate drawings, find details, plan nesting, export takeoffs and track revisions in one workspace.
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
Published · Updated
Alkali has moved from an AI process-engineering thesis into a concrete steel-estimating workflow.
What it does
The current Alkali homepage focuses on structural-steel estimation. It describes AI takeoffs for beams, columns, joists, braces and base plates; built-in AISC weights; nesting plans; exports to Tekla, Bluebeam, Adobe and Excel; revision detection; and a shared estimate workspace.
That is more specific than the earlier YC launch description, which framed Alkali as ProcessMate for chemical-process design and equipment sourcing. The current public product is steel estimation. I would use the current homepage for buyer fit and keep the earlier chemical-engineering story as company history rather than blending the two products.
Why I’d look closer
The product is aimed at a real estimating bottleneck: turning drawings and revisions into a shared, auditable material view. The homepage says it has been used across 2,000+ steel projects and 48 states, and that the anonymized subset represents more than 10% of U.S. steel. Those are company-published claims, not an independent market measurement.
The founder background fits the domain. The YC launch describes Emmett Goodman’s chemical-engineering and AWS GenAI Labs experience and Victor Miller’s chemical engineering, laboratory and data-science background. The company is applying technical training to a domain where a missed shape, length or revision can become cost and rework.
What I’d ask
I would test one real bid set: count accuracy, revision handling, nesting assumptions, export quality and the review path when the drawing is ambiguous. Pricing is not published in the checked pages. The homepage’s “no signup required” path is useful for exploration, but a fabricator still needs to understand data retention, project collaboration and who owns the final estimate.
My editorial take
Shortlist Alkali for a steel fabricator or estimator who wants to cut takeoff work without abandoning the tools already in the shop. It is less relevant if your work is not structural steel or if your bottleneck is procurement rather than estimating. The product earns attention because the workflow is concrete: drawing in, auditable estimate out, revision tracked.
Quick facts
| Field | Sourced detail |
|---|---|
| Product | AI steel takeoff, nesting, search and revision workspace |
| Buyer | Structural-steel fabricators and estimating teams |
| Public scale claim | 2,000+ projects and 48 states, company-published |
| Pricing | Not published in the checked pages |
| Main question | Can the output survive a real bid-set review and export process? |
Sources checked
| Source | Checked |
|---|---|
| YC company profile | 2026-09-19 |
| Alkali homepage | 2026-09-19 |
| Alkali blog | 2026-09-19 |
Cohort context
Alkali is listed in Spring 2025. In our 2026-09-18 directory snapshot, 97 of 143 listed companies in that cohort have YC’s primary industry label B2B (67.8%). This is a current-directory comparison, not an original intake count or a performance ranking. Nine-cohort dataset.
Public website snapshot
Observed 2026-09-19T16:15:27.419Z in raw homepage HTML. This records visible metadata and advertised links, not agent execution or product quality.
| Signal | Homepage observation |
|---|---|
| Product description metadata | Observed |
| Canonical link | Not observed in this response |
| H1 or H2 heading | Observed |
| Typed structured data | Not observed in this response |
| Docs/developer link | Not observed in this response |
| Pricing link | Not observed in this response |
| llms.txt link | Not observed in this response |
| Markdown alternate | Not observed in this response |
Public observations · Collection method. Missing links here do not establish that a capability or file is absent elsewhere.
About the author
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
I cofound Lazyweb and publish Mudpie. This is an owner-written publication, not an independent testing organization. Research notes distinguish observations, sourced reporting and editorial judgment.
First1000 ↗ · X ↗