Means & Methods Lab, home
Facts: desk research from vendor sites · Reviews: members' opinions

Davis AI vs Marengo

Two firms that sell AI-assisted design work as a service rather than software, aimed at different buildings, so this is partly like-for-like (our reading). Davis AI is a Paris-based AI-assisted architecture firm that delivers feasibility studies and architectural designs for real estate developers and investors. Marengo is an AI-assisted engineering firm that does data center pre-construction engineering, from site due diligence through feasibility, concept design and FEED / permitting design. Neither publishes prices. Desk research from vendor sites read Nov 8, 2026; no winner.

DESK RESEARCH FROM VENDOR SITES. Compiled from vendors' public websites (accessed Davis AI Nov 19, 2026; Marengo Nov 18, 2026). Vendor claims are not verified by Means & Methods Lab. No scores or rankings.

In short: Davis AI (AI feasibility & design service for real estate; no public prices (quote / sales)) vs Marengo (AI-assisted data center engineering firm; no public prices (quote / sales)). Facts last checked Davis AI Nov 19, 2026; Marengo Nov 18, 2026. No winner declared; desk research from vendor sites.

Side by side

Davis AI vs Marengo: vendor-published facts side by side
Sheet rowDavis AIMarengo
CategoryAI feasibility & design service for real estateAI-assisted data center engineering firm
AI role (our label)AI-firstAI-first
What it doesParis-based AI-assisted architecture firm that says it delivers feasibility studies and architectural designs in days, not months, pairing AI with architect judgment at each stage. It says it does not sell software but delivers results. The homepage says Davis ingests a client's data rooms, enriches them with regulatory, market and site data gathered by its own agents, turns that into regulatory, financial and architectural constraints, and has its team validate each stage. Listed services: feasibility studies, regulatory analysis, investment analysis, volumetrics, floorplans, unit-mix optimization and space planning. A May 1, 2026 technology post introduces Gaudi-1, a proprietary discrete-diffusion floor-plan generation model, says its in-house architects stay in the loop, and claims state-of-the-art results on the RPLAN and MSD benchmarks (vendor claim, not verified). The site announces a $5.5M raise.An engineering firm rather than a software product: its engineering team uses in-house AI tools to speed up data center pre-construction engineering, from site due diligence through feasibility, concept design and FEED / permitting design. The site lists deliverables including a site constraints register, site risk assessment, developable-area analysis, utility and infrastructure assessment, test fit, master plan, decision-ready feasibility report, basis of design and design criteria, CAD layouts, electrical architecture and single-line diagrams, cooling, civil and structural concepts, a 3D coordinated model and permit-design documentation, and contrasts multi-objective optimization with manual trade-off assessment. It says designs are reviewed and validated by licensed engineers before delivery, client IP is never used to train models, and the client owns everything Marengo produces. Engagements start with one site, benchmarked against your conventional workflow, before scaling. Y Combinator batch: Summer 2026; its YC profile describes it as an AI-native engineering firm automating engineering design, starting with data centers, and the site names Y Combinator and a16z Scout Fund as backers while the profile names YC, Exor and Future Planet Capital (company statements). Design-cycle, feasibility time and cost, and concepts-explored comparisons are vendor claims, and the YC profile's revenue and contract figures are not used.
Who it's forReal estate developers and investors evaluating sites and early designs.Data center developers and owners evaluating sites and campuses.
Pricing (as publicly stated)No public prices: the vendor site has no pricing or plans link in its navigation or footer, /pricing and /plans return a not-found page, and the call to action is Book a demo. Confirm current pricing with the vendor.No public prices: Marengo sells engineering services; the call to action is Give Us a Site and marengox.com/pricing returns a not-found page. Confirm with Marengo.
User reviewsNo reviews yet. Read or writeNo reviews yet. Read or write
Vendor links
Sources

Facts last checked: Davis AI Nov 19, 2026; Marengo Nov 18, 2026. Prices change; check the vendor page before relying on them.

Where the vendor descriptions differ

  • Partly like-for-like: both say they deliver results rather than selling software, and both keep people in the loop. Davis says its team validates each stage and its in-house architects stay in the loop; Marengo says designs are reviewed and validated by licensed engineers before delivery. The building types differ: Davis lists feasibility, unit mix and floorplans for real estate development, and Marengo is built around data center sites and campuses.
  • What you get: Davis lists feasibility studies, regulatory analysis, investment analysis, volumetrics, floorplans, unit-mix optimization and space planning. Marengo lists a site constraints register, site risk assessment, developable-area analysis, utility and infrastructure assessment, test fit, master plan, feasibility report, basis of design, CAD layouts, electrical architecture and single-line diagrams, cooling, civil and structural concepts, a 3D coordinated model and permit-design documentation.
  • Inputs: Davis says it ingests a client's data rooms and enriches them with regulatory, market and site data gathered by its own agents, then turns that into regulatory, financial and architectural constraints. Marengo starts from a site (its call to action is Give Us a Site) and says engagements start with one site, benchmarked against your conventional workflow, before scaling.
  • AI: Davis describes Gaudi-1, its proprietary floor-plan generation model, in a May 1, 2026 technology post (its benchmark claims are vendor claims, not verified here). Marengo describes in-house AI tools used by its own engineering team and contrasts multi-objective optimization with manual trade-off assessment.
  • Data and IP: Marengo says client IP is never used to train models and the client owns everything it produces. The Davis pages we read do not state an IP or training policy, so ask.
  • Pricing (Marengo read Nov 8, 2026; Davis AI from its Nov 8, 2026 entry): neither lists dollar amounts. Davis has no pricing link, /pricing and /plans return a not-found page, and its call to action is Book a demo. Marengo's marengox.com/pricing returns a not-found page and its call to action is Give Us a Site. See each tool page for detail.
  • Not used here: Davis speed lines and funding amount; Marengo design-cycle, feasibility time and cost figures and YC revenue and contract figures (vendor material). Facts for both tools come from vendor websites and Marengo's Y Combinator profile (desk research).

Paraphrased from the sources listed above. Vendor claims are reported as claims.

What users say

The facts above come from vendor websites. Whether either tool is any good for your work is something only people who use them can say. These are member reviews, all ratings shown, never edited or removed for being negative.

Used either of these? Share your take

Good or bad, an honest review from an estimator, PM or super helps the next team more than another vendor demo. You post under a handle; your email is never shown.

Questions to ask in your review

  • Whether your project is a data center campus (Marengo's focus) or a residential or mixed-use development where unit mix and floorplans matter (Davis AI's listed services)
  • Which licensed architects or engineers sign each firm's deliverables in your jurisdiction
  • What each would charge for one site
  • What did you use it for, on what kind of work and project size?
  • How long did setup and training take, and who on your team did it?
  • What does it cost you in practice, compared with what the sales call said?

How reviews work: review guidelines and methodology. All reviews: reviews hub.

Related