Hestus
AI copilot for mechanical engineers, integrating with CAD software to automate repetitive design work.
Website: https://www.hestus.co/
Cover Block
Publicly reported
| Field | Detail |
|---|---|
| Name | Hestus |
| Tagline | AI copilot for mechanical engineers, integrating with CAD software to automate repetitive design work [Hestus, retrieved 2024] |
| Headquarters | San Mateo, California [San Francisco Business Times, January 2025] |
| Founded | 2024 [San Francisco Business Times, January 2025] |
| Stage | Seed [Unite.AI, October 2026] |
| Business Model | SaaS |
| Industry | Deeptech |
| Technology | AI / Machine Learning [Hestus, retrieved 2024] |
| Geography | North America |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (2): Sohrab Haghighat and Kevin Chu [San Francisco Business Times, January 2025] |
| Funding Label | Seed, total disclosed $8.9 million [Unite.AI, October 2026] |
Links
Publicly reported
- Website: https://www.hestus.co/
- Y Combinator: https://www.ycombinator.com/companies/hestus-inc
- Careers: https://www.ycombinator.com/companies/hestus-inc/jobs/nmBIPkW-machine-learning-engineer
Summary and Signal
PUBLIC Hestus builds an AI copilot for mechanical engineers inside CAD software, and it merits investor attention now because it sits at a credible intersection of applied AI, design workflow automation, and fresh financing, with a reported $7.4 million seed round in October 2026 bringing total disclosed funding to $8.9 million [Unite.AI, October 2026] [San Francisco Business Times, January 2025]. The company was founded in 2024 in San Mateo by Sohrab Haghighat and Kevin Chu, two former Cruise engineers, and entered the market with a narrow wedge rather than a broad platform claim: Sketch Helper for Autodesk Fusion 360, released in September 2024 according to SpaceNews [SpaceNews, January 2025] [San Francisco Business Times, January 2025].
The product thesis is straightforward. Hestus aims to automate repetitive mechanical design work by turning design intent into sketches, constraints, and dimensions, while adding autocomplete-style suggestions inside an existing CAD environment rather than asking users to adopt a net-new system [Hestus, retrieved 2024] [SpaceNews, January 2025]. That integration-first approach matters because CAD behavior is sticky, and the practical question is less whether engineers want AI assistance than whether they will trust it inside production workflows where design errors carry downstream manufacturing cost [Make:, retrieved 2026] [Unite.AI, October 2026].
The team has relevant domain credibility for that bet. Haghighat previously served as a senior engineering manager for planning and controls at Cruise and earlier founded SpaceRyde, while Chu was a director of autonomous engineering at Cruise and was reported to be among the company's first 10 employees [TechCrunch, June 2019] [Y Combinator] [SpaceNews, January 2025]. Investor support also appears to have matured quickly, from a $1.5 million pre-seed led by Liquid2 in January 2025 to an Insight Partners-led seed round in October 2026, with Y Combinator, Rock Yard Ventures, Chalfen Ventures, Garage Capital, Kyle Vogt, Eric Migicovsky, and Benjamin Bryant among the disclosed backers [San Francisco Business Times, January 2025] [R&D World, January 2025] [Unite.AI, October 2026].
The business model is framed as SaaS, but the public record is still light on pricing, retention, and enterprise deployment depth. The main public traction signal is a reported base of more than 5,000 engineers using Hestus tools, alongside a company-adjacent claim that the user base doubled since January 2026; useful signals, but not yet a substitute for customer concentration, paid conversion, or renewal data [Unite.AI, October 2026] [LinkedIn, 2026]. Over the next 12 to 18 months, the points to watch are whether Hestus expands beyond Fusion 360 into SolidWorks as indicated in public materials, converts usage into repeatable paid adoption, and shows that AI assistance in CAD can become part of a durable engineering workflow rather than a useful point tool [Make:, retrieved 2026] [LinkedIn, 2026] [Hestus, retrieved 2024].
One source, partially checked -- Relies on named public sources including San Francisco Business Times, SpaceNews, R&D World, Unite.AI, Hestus, and Y Combinator, with some product and traction details partially corroborated and some recent adoption claims sourced to company-adjacent posts.
Taxonomy Snapshot
| Axis | Value |
|---|---|
| Stage | Seed |
| Business Model | SaaS |
| Industry / Vertical | Deeptech |
| Technology Type | AI / Machine Learning |
| Geography | North America |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (2) |
| Funding | $8,900,000 disclosed |
Company Overview
PUBLIC
Hestus is a San Mateo, California startup building software for mechanical engineers, with an early wedge inside Autodesk Fusion rather than a standalone design system [Hestus, retrieved 2024]. Public company profiles describe it as founded in 2024, with Sohrab Haghighat listed as co-founder and CEO and Kevin Chu listed as co-founder and CTO [Crunchbase] [Y Combinator].
The public record is still thin, but the chronology is relatively clear. Hestus launched its first product presence around Autodesk Fusion in 2024 through its website positioning as "Autocomplete for CAD," and Crunchbase reflects a 2024 founding date [Hestus, retrieved 2024] [Crunchbase]. By 2026, the company was presenting itself through the Y Combinator directory as Hestus, Inc., part of the S24 batch, which helps anchor both the legal naming convention and accelerator affiliation in the absence of state filing detail in the sourced materials [Y Combinator].
A second milestone is financing cadence rather than product breadth. Crunchbase records a seed financing profile for Hestus, while the company website continues to frame the business around Autodesk Fusion-based workflow assistance, suggesting the company remains in the focused product-build stage rather than broad platform expansion [Crunchbase] [Hestus, retrieved 2024]. That combination, a recent founding, a narrow initial integration surface, and a visible startup formation path through YC, is consistent with an early enterprise software company still defining its durable distribution channel [Y Combinator] [Hestus, retrieved 2024].
One source, partially checked -- Confirmed primarily by the company website, Crunchbase, and Y Combinator; headquarters, founding year, leadership, and YC affiliation are corroborated, but state filing detail was not present in the sourced public record.
The Product and the Stack
Public record plus analysis
Hestus is approaching a familiar engineering bottleneck from a specific angle: the repetitive sketching work that sits upstream of downstream design errors. Public materials describe the product as an AI copilot for mechanical engineers, initially delivered inside Autodesk Fusion 360 as "Sketch Helper," where it translates design intent into sketches, constraints, and dimensions and offers autocomplete-style suggestions as a user works [Hestus, retrieved 2024] [SpaceNews, January 2025] [Make:, retrieved 2026]. The company website is more concrete than most early product pages, stating that the tool is available for Autodesk Fusion on Mac and Windows and showing a workflow where Hestus reads an in-progress sketch and proposes geometry such as centerlines with a single keypress [Hestus, retrieved 2024].
The evidence suggests Hestus is not trying to replace incumbent CAD systems so much as sit inside them and compress low-value design iteration. That interpretation is consistent across the website, press coverage, and a hands-on writeup from Make:, which describes Hestus as a plugin for Autodesk Fusion with SolidWorks support coming in June [Hestus, retrieved 2024] [San Francisco Business Times, January 2025] [Make:, retrieved 2026]. A LinkedIn post cited in the source set also says Hestus is rolling out on SolidWorks, but that claim is weaker than the Fusion evidence and should be read as directional rather than fully verified product availability [LinkedIn, 2026].
There is very little public technical-stack disclosure, so the safer read is product-surface first, infrastructure second. The company and third-party coverage consistently frame the core capability as AI recommendations delivered during the design process, with the claimed outcome of fewer redesign cycles, less trial and error, and fewer manufacturing mistakes, but these remain benefit claims rather than independently measured performance results [Unite.AI, October 2026] [SpaceNews, January 2025] [Hestus, retrieved 2024]. Tech-stack details are limited; the only defensible inference is that machine learning is a material internal competency given open roles for machine learning engineers (inferred from job postings) [Y Combinator] [Y Combinator, retrieved 2026].
One source, partially checked -- Product positioning and Fusion 360 integration are corroborated by the company website, SpaceNews, and Make:, while SolidWorks availability and deeper technical implementation details remain only partially corroborated.
The Market They Are Entering
PUBLIC
This market matters now because mechanical design remains a large, labor-intensive workflow inside industrial software, while recent AI tooling is starting to move from text and code into domain-specific engineering tasks, a shift that creates room for application-layer companies such as Hestus if they can fit into existing CAD habits rather than asking engineers to replace them outright [San Francisco Business Times, January 2025] [SpaceNews, January 2025] [Unite.AI, October 2026].
The public record here is thinner than an institutional investor would want for formal TAM work. No cited third-party market study in the source set gives a direct size for AI copilots in mechanical CAD, so any market framing has to stay one step back and rely on adjacent demand signals. What is corroborated is the workflow wedge: Hestus is positioning around repetitive sketching, constraints, dimensions, and design-intent translation inside Autodesk Fusion 360, with additional CAD support reported as planned or rolling out [Hestus] [SpaceNews, January 2025] [Make:] [LinkedIn, 2026]. That places the company at the intersection of established CAD software spend and newer AI productivity spend, which is narrower than the full engineering software stack but still meaningful if adoption expands from a plugin into a daily workflow layer.
Demand drivers in the sources are practical rather than thematic. Multiple reports describe the product as aiming to reduce redesign cycles, trial and error, and manufacturing mistakes, which suggests the value proposition is tied to engineering throughput and fewer downstream errors rather than abstract "AI transformation" budgets [SpaceNews, January 2025] [R&D World, January 2025] [Hestus]. The fact pattern also suggests some early receptivity from practitioners: Unite.AI reported more than 5,000 engineers using Hestus tools as of October 2026, though that figure appears to trace back to company claims and should be treated as directional rather than independently audited [Unite.AI, October 2026].
The adjacent markets are broader than mechanical CAD alone. Fusion 360 is Hestus's initial surface, but the cited product descriptions point toward a wider category that includes product lifecycle design workflows, simulation-adjacent design assistance, and engineering productivity software used by teams building drones, robots, industrial equipment, and consumer products [SpaceNews, January 2025] [Unite.AI, October 2026]. Substitute behavior also matters. Engineers can continue using native CAD features, macros, templates, or human review loops instead of an AI copilot, so Hestus is competing not only with software vendors but with existing workflow inertia.
Regulatory and macro forces are present, even if the current sources do not tie them directly to Hestus's revenue motion. Manufacturing and hardware teams are under persistent pressure to shorten design cycles and avoid costly iteration, especially in sectors where physical prototyping is expensive and supply chains remain sensitive to delays [SpaceNews, January 2025]. At the same time, engineering software buyers tend to be conservative because design errors propagate into production, which means any AI assistant in this category will likely face a higher trust threshold than general-purpose workplace copilots. That cuts both ways: slower adoption can lengthen sales cycles, but if the software becomes reliable enough to sit inside core design workflows, switching friction can also rise.
| Market lens | Public evidence | Interpretation |
|---|---|---|
| Core wedge | AI copilot inside Autodesk Fusion that autocompletes sketches, constraints, and dimensions [Hestus] | Starts as workflow augmentation inside existing CAD rather than a full CAD replacement |
| Initial user domain | Mechanical engineers and hardware teams building physical products [San Francisco Business Times, January 2025] [Unite.AI, October 2026] | The addressable user base is tied to industrial and hardware design, not general knowledge work |
| Platform expansion | SolidWorks support reported as planned or rolling out [Make:] [LinkedIn, 2026] | Category upside depends in part on expanding beyond a single CAD ecosystem |
| Demand proof point | More than 5,000 engineers reportedly use Hestus tools [Unite.AI, October 2026] | Suggests early market pull, but the figure remains company-sourced in public reporting |
The evidence supports a real workflow pain point and a plausible category opening, but it does not yet support a hard market-size model from named third-party research. For now, the cleaner public read is that Hestus is entering a narrow but potentially expandable slice of engineering software where adoption will depend more on accuracy, trust, and CAD-platform reach than on headline AI budget growth.
One source, partially checked -- Section relies on named public reporting and the company website for workflow positioning, but no third-party market-sizing report is present in the source set, and the main adoption figure appears company-sourced through media coverage.
The Competitive Field
MIXED Hestus is positioning itself less as a replacement for CAD systems than as an AI workflow layer inside them, which places it in a narrower but potentially useful slot between incumbent design platforms and general-purpose engineering productivity tools [Hestus, retrieved 2024] [San Francisco Business Times, January 2025] [SpaceNews, January 2025].
The competitive map is defined first by incumbents, even though the available source set does not name direct startup rivals. Autodesk Fusion 360 is the current host platform for Hestus's first product, Sketch Helper, which means Autodesk is both an enabling channel and the baseline alternative, because any native improvement in sketch automation or constraint assistance could absorb part of Hestus's value proposition [Hestus, retrieved 2024] [SpaceNews, January 2025]. SolidWorks sits in the next important segment as another major CAD environment Hestus has said it plans to support or is rolling out against, according to public reporting and secondary public posts, which suggests the company is following installed engineer workflows rather than asking teams to switch core design systems [Make:, retrieved 2026] [LinkedIn, 2026].
The adjacent substitute set is broader than it first appears. Mechanical engineers can still solve repetitive sketching and cleanup work with manual CAD operations, internal templates, macros, and established team conventions, and those substitutes usually come bundled into existing software budgets rather than requiring a separate software decision. That matters because Hestus's public pitch is strongest where the design task is repetitive and sketch-heavy, but the current evidence base does not yet show a broader platform claim across simulation, PLM, or downstream manufacturing software categories [Hestus, retrieved 2024] [SpaceNews, January 2025].
Where Hestus appears strongest today is talent and workflow fit. The founders' prior engineering leadership at Cruise gives the company technical credibility with complex physical-product teams, and the product is framed around concrete sketch actions such as adding constraints, dimensions, and cleanup rather than a broad AI assistant concept, which makes the initial wedge easier to evaluate in production design work [San Francisco Business Times, January 2025] [SpaceNews, January 2025] [Crunchbase]. The edge is real but perishable. Because the product is currently described as a plugin or layer on top of Autodesk Fusion, its distribution still depends on user adoption inside a third-party ecosystem that Hestus does not control, and a host platform can always narrow the gap if it chooses to ship similar native features [Hestus, retrieved 2024] [Make:, retrieved 2026].
The company's main exposure is therefore platform dependence rather than capital scarcity alone. Insight Partners, Liquid2, Y Combinator, and other backers give Hestus more room to hire and ship than a bootstrapped tools vendor would have, but capital does not remove the structural risk that Autodesk or another major CAD vendor owns the primary user relationship, the default interface, and the procurement path for larger accounts [Unite.AI, October 2026] [San Francisco Business Times, January 2025] [Y Combinator]. A second exposure is scope. Public materials support an early lead around sketch assistance inside Fusion 360, yet they do not establish that Hestus already owns higher-value layers such as enterprise administration, systems-of-record data, or deeply embedded simulation workflows, where incumbent suites tend to be harder to displace [Hestus, retrieved 2024] [SpaceNews, January 2025].
Over the next 18 months, the most plausible competitive scenario is not a clean winner-take-all outcome but a race to become the default AI assist layer for mechanical design inside existing CAD seats. Autodesk is the likely winner if native CAD copilots become a standard feature expectation, because it already controls the core product environment and can distribute new functionality into an installed base at low marginal acquisition cost [Hestus, retrieved 2024]. Hestus is the likely winner if engineers continue to prefer specialized assistance that works inside incumbent tools and if the company can extend from Fusion 360 into SolidWorks without losing product quality, because that would turn a single-plugin wedge into a cross-platform workflow product [Make:, retrieved 2026] [LinkedIn, 2026]. In the same scenario, Hestus is also the most plausible loser if host platforms close the feature gap before the company establishes durable data advantages or an enterprise foothold beyond individual user productivity [Hestus, retrieved 2024] [Unite.AI, October 2026].
One source, partially checked -- This section relies on named public sources for Hestus's product position, platform integrations, funding, and founder backgrounds, but the available record does not identify named startup competitors and parts of the SolidWorks rollout narrative rely on single-source or company-adjacent evidence [Hestus, retrieved 2024] [SpaceNews, January 2025] [San Francisco Business Times, January 2025] [Unite.AI, October 2026] [Make:, retrieved 2026] [LinkedIn, 2026].
Opportunity
PUBLIC
The prize here is large if Hestus can become the software layer that sits between mechanical engineers and the CAD systems they already use, because that would place it inside a daily workflow that is expensive to change and central to how physical products get designed [Hestus, retrieved 2024] [San Francisco Business Times, January 2025].
The headline opportunity is not simply to sell a helpful plugin. It is to become the default intelligence layer for mechanical design work across incumbent CAD environments, starting with Autodesk Fusion 360 and then extending into other systems such as SolidWorks that the company has said are in progress or rolling out [Hestus, retrieved 2024] [SpaceNews, January 2025] [Make:, retrieved 2026] [LinkedIn, 2026]. That outcome is reachable, not merely aspirational, because Hestus already appears to have chosen a practical wedge: automate repetitive sketching, constraints, and dimensions inside an existing design tool rather than asking engineers to switch systems wholesale [Hestus, retrieved 2024] [SpaceNews, January 2025]. The public record also suggests some early usage density, with Unite.AI reporting that more than 5,000 engineers use Hestus tools and the company raising an Insight Partners-led $7.4 million seed in October 2026 after an earlier $1.5 million financing led by Liquid2 and Rock Yard Ventures [Unite.AI, October 2026] [San Francisco Business Times, January 2025].
A few upside paths stand out from the available evidence.
| Scenario | What happens | Catalyst | Why it's plausible |
|---|---|---|---|
| Fusion wedge becomes system of record for AI assistance | Hestus becomes the standard add-on for Autodesk Fusion users who want faster sketching and fewer repetitive inputs | Continued product improvement inside Fusion plus visible time-savings claims on core tasks [Hestus, retrieved 2024] | The product is already live in Fusion, available on Mac and Windows, and described as finishing constraints, dimensions, and cleanup within the sketch workflow [Hestus, retrieved 2024] [SpaceNews, January 2025] |
| Multi-CAD expansion broadens the addressable base | Hestus extends from Fusion into SolidWorks and potentially other CAD environments, turning a narrow wedge into a broader platform | Successful SolidWorks rollout or broader cross-platform support [Make:, retrieved 2026] [LinkedIn, 2026] | Public sources consistently describe Fusion as the first surface rather than the only one, with SolidWorks support planned, coming, or rolling out [SpaceNews, January 2025] [Make:, retrieved 2026] [LinkedIn, 2026] |
| Workflow intelligence compounds into enterprise standardization | Hestus moves from single-user productivity into team-level design standardization and error reduction for hardware companies | Evidence that recommendations across the design process reduce redesign cycles and manufacturing mistakes [Unite.AI, October 2026] [LinkedIn, 2026] | The company is positioned around repetitive engineering work and design recommendations, which are categories that can expand from individual assistance into process control if accuracy holds up in production use [Unite.AI, October 2026] [Hestus, retrieved 2024] |
The compounding logic is straightforward. If Hestus saves time on the first sketching tasks, it earns more usage; if it earns more usage, it should collect more examples of how engineers express design intent inside CAD; and if those examples improve recommendations, the product becomes more useful and harder to displace inside the day-to-day workflow. Public evidence only shows the first steps of that loop, but those steps are real enough to matter: the homepage centers on autocomplete inside Autodesk Fusion, SpaceNews reported a September 2024 release for Sketch Helper, and Unite.AI later reported more than 5,000 engineers using Hestus tools [Hestus, retrieved 2024] [SpaceNews, January 2025] [Unite.AI, October 2026].
There is also a distribution advantage in the company’s current posture. By integrating with software engineers already use, Hestus can ride incumbent CAD adoption rather than fight it, which is usually the lower-friction route in technical workflows [San Francisco Business Times, January 2025] [Hestus, retrieved 2024]. The founder profile helps the case at the margin: Sohrab Haghighat and Kevin Chu both came from Cruise, with Haghighat previously founding SpaceRyde and Chu serving in autonomous engineering leadership, which supports the view that the team understands complex engineering environments rather than approaching the category as a generic software play [SpaceNews, January 2025] [TechCrunch, June 2019] [Y Combinator].
The size of the win is harder to quantify cleanly because the source set here does not include a named market-size study or a tight public comparable for AI-native CAD assistance. Even so, the upper-bound logic is visible. If Hestus becomes a cross-platform intelligence layer used by a meaningful share of professional mechanical engineers, then the outcome is larger than a point solution and closer to a durable design-software platform, which could justify multi-billion-dollar equity value under a strong software multiple and broad workflow ownership (scenario, not a forecast). That scenario rests on public evidence that the company has already launched in Fusion, is extending toward SolidWorks, has reported thousands of users, and has attracted seed backing from Insight Partners, Y Combinator, Liquid2, Rock Yard Ventures, Chalfen Ventures, and Garage Capital [Hestus, retrieved 2024] [Unite.AI, October 2026] [San Francisco Business Times, January 2025] [Y Combinator] [Crunchbase].
One source, partially checked -- This section relies on a mix of company materials, named news coverage, Crunchbase, Y Combinator, and LinkedIn posts; the core product and funding claims are partially corroborated, while adoption scale and some platform-expansion signals remain single-source or company-adjacent.
Sources
Publicly reported
[Hestus, retrieved 2024] Hestus | Autocomplete for CAD | https://www.hestus.co/
[San Francisco Business Times, January 2025] Former Cruise engineers launch Hestus, an AI-powered CAD software startup | https://www.bizjournals.com/sanfrancisco/news/2025/01/30/hestus-ai-cad-software-startup-san-mateo.html
[Unite.AI, October 2026] Hestus Raises $7.4 Million Seed to Bring AI Into Mechanical Engineering | https://www.unite.ai/hestus-raises-7-4-million-seed-to-bring-ai-into-mechanical-engineering/
[SpaceNews, January 2025] SpaceRyde founder’s startup offers CAD copilot | https://spacenews.com/spaceryde-founders-startup-offers-cad-copilot/
[Make:, retrieved 2026] Hands-On With Hestus Autocomplete for CAD | https://makezine.com/article/digital-fabrication/cad/hands-on-with-hestus-autocomplete-for-cad/
[TechCrunch, June 2019] SpaceRyde wants to make access to space more available and more affordable | https://techcrunch.com/2019/06/24/spaceryde-wants-to-make-access-to-space-more-available-and-more-affordable/
[Y Combinator] Hestus, Inc.: An AI-powered CAD software | https://www.ycombinator.com/companies/hestus-inc
[R&D World, January 2025] Hestus secures $1.5M in seed funding for AI-backed CAD | https://www.rdworldonline.com/hestus-secures-1-5m-seed-funding-to-streamline-cad-workflows-with-targeted-ai/
[LinkedIn, 2026] Femi Palmer - Managing Partner at Initio Capital | Ex-Bain | https://www.linkedin.com/in/olufemi-palmer/
[LinkedIn, 2026] Arash Naghib Lahouti - Sr. Product Design Engineer (Fluid ...) | https://www.linkedin.com/in/arash-naghib-lahouti-56266815/
[Crunchbase] Seed Round - Hestus - Crunchbase Funding Round Profile | https://www.crunchbase.com/funding_round/hestus-seed--5b5b2c4c
Articles about Hestus
- Hestus Puts Autocomplete Into the Hands of 5,000 Mechanical Engineers — The Y Combinator-backed startup, founded by former Cruise leaders, raised $7.4 million to build an AI copilot for CAD software like Fusion 360 and SolidWorks.