Startup from MIT · Company Brain for Engineering Design

Engineering knowledge,
wired to your geometry.

Metis is engineering design intelligence for engineers — surfacing DFM guidelines, lessons learned, and design standards in real time, connected directly to the part you are working on.

Request a Demo See How It Works
Faster design cycle
with fewer errors
The right standards and lessons learned surface while engineers design, catching issues before costly late-stage changes.
Accumulate knowledge
before experts leave
Tacit know-how from your most experienced engineers is captured and structured, so it stays with the company.
The Problem

The knowledge is there.
Finding it is the problem.

Design Release Engineers in Automobile, Consumer Electronics, Aerospace, etc. spend a significant part of every day on indirect work — hunting across PLM systems, shared drives, legacy PDFs, and tribal memory for knowledge that already exists somewhere in the organization, while spinning their CAD to understand which information is relevant to their parts.

The cost is not just time. For example, a single late-stage design change in automotive can cost millions in tooling and program delay. When institutional knowledge walks out the door, those costs multiply across every future program.

50%
of engineering time is spent on non-value-added work — with half of engineering leaders estimating the true figure is 20–39% of the workday. (CoLab, 2024) This increases to 50% when we include meetings.
$M per change
A single late-stage design change in automotive can cost millions in tooling rework and program delay. The knowledge gap is structural — and worsens as workforces turn over.
The Product

The company brain that turns
company knowledge, tacit and explicit,
into structured knowledge.

Metis integrates directly into and around CAD, enabling engineers to retrieve general and company-specific engineering knowledge via natural language. It unifies text, image, and geometry information, providing actionable design insights in context. Engineers spend less time searching and more time on actual engineering design.

The system combines a general engineering layer (industry standards, DFM rules, material properties) with a customer-specific layer (company lessons learned, internal guidelines, proprietary standards) to build the company's collective intelligence.

L1
Just-in-Time Collective Intelligence
Cursor-position spatial context + retrieval system surfaces DFM guidelines, lessons learned, and design standards in real time.
L2
Tacit to Structured Knowledge
The system probes what it does not yet know, drawing out the tacit knowledge that lives only in engineers' heads.
The Vision

Every engineering decision
is informed by everything
your organization knows.

Every engineer accumulates judgment that never makes it into a document — the reason a fillet radius got bumped up after a field failure, the supplier quirk that changes a tolerance, the fix that only lives in one person's head. Metis captures that tacit knowledge directly from engineers, including your most experienced ones, as they work, and turns it into structured, retrievable knowledge the moment it's created.

Over time, this becomes an institutional memory that compounds — accumulating with every design decision and passing that expertise on to the next generation of engineers, so it stays with the company even as people move on.

The Team

Built by engineers who lived the problem,
with MIT research.

Headshot of Hiroki Ogasawara
Hiroki (Hiro) Ogasawara
Co-Founder & CEO

10+ years as a Design Release Engineer at Toyota Motor Corporation — Assistant Manager, Body Design Engineering, designing vehicles such as the RAV4, Highlander, and Sienna in both Japan and the United States. 14 patents and equivalent rights in automobile design. He earned a dual master’s degree in Systems Design and Management (SDM) and Electrical Engineering and Computer Science (EECS) from MIT. While at MIT, he participated in Ford’s Strategy Leadership Program, where he helped formulate strategies for next-generation electric vehicles, and in Apple’s Engineering Leadership Program, where he worked as a product designer on the planning and design of next-generation products. He received the 2024 MIT SDM Award for Leadership, Innovation, and Systems Thinking.

Toyota, Ford & Apple · 10 yrs Design Engineer CAD Expert MIT SDM + EECS
LinkedIn →
Headshot of Rowell Castro
Rowell Castro
Co-Founder & CTO

9+ years at Shell, driving internal digital transformation as the Digital Innovations Manager for the Philippines, Singapore, Malaysia, Thailand, and the UAE. He has twice received the Shell VP Award. He earned a dual master’s degree in Systems Design and Management (SDM) and Electrical Engineering and Computer Science (EECS) from MIT. He is affiliated with the MIT Computer Science and Artificial Intelligence Laboratory. His MIT thesis, “Semantic Learning in Engineering Design,” forms the core technology of the solution. He has experience founding a startup in the Philippines.

Shell · 9 yrs Digital Innovations Manager AI / ML 2× Shell VP Award MIT SDM + EECS
LinkedIn →
Traction

From discovery to demonstration
— the feedback is consistent.

🎙
100+ customer discovery interviews
Completed across automobile companies and suppliers, validating the core problem and shaping product decisions.
⚙️
Working prototypes
Streamlining engineering workflows to address pain points current engineers have.
🏭
OEMs and suppliers engagement (Paid POC)
Active interest from engineers and a director at automobile companies and suppliers, following prototype demonstrations. Engineering and leadership stakeholders confirmed the problem and we landed the first paid POC.
🎓
Startup from MIT
Institutional backing from MIT Sandbox Innovation Fund Program. Originated in Professor Bill Aulet's Entrepreneurship 101. One of the top 7 out of 80 MIT startups in MIT $100K Entrepreneurship Competition (Link to the pitch).
🚀
Top 10% of Y Combinator applicants
Ranked in the top 10% of applicants for Y Combinator Summer 2026 on our first attempt.
News

Latest news

Metis selected for UTokyo IPC's 1stRound program
UTokyo IPC has selected Metis for full selection in the 15th cohort of 1stRound, one of four teams out of eight selected.
Read more →
Metis selected for Founder University Japan's US Immersion Week
Metis is one of 10 companies selected from Founder University Japan, the JETRO × LAUNCH pre-accelerator, for US Immersion Week in San Francisco, October 19–23.
Read more →
Metis named a finalist in the 2026 MIT $100K Entrepreneurship Competition
Metis, then competing as DesignFlowAI, was one of seven finalists in MIT's flagship startup competition, selected through three rounds from about 80 MIT startups.
Read more →
Get In Touch

Request a demo
or discuss a pilot.

We are actively running pilots with design engineering teams in industries. If you are a design engineer, engineering director, or investor — we want to hear from you.

Pilot Inquiries Investor Relations Engineering Roles
We respond within one business day.

Request received.

Thanks for reaching out — we will be in touch within one business day.