Mariana Silva

Mariana Silva

I'm a product leader passionate about making complex technology feel simple. I enjoy partnering with researchers, engineers, and customers to turn ambitious technical ideas into intuitive products that solve meaningful problems and scale.

.about

I came to product from an engineering background. I studied mechanical engineering, started my career building machine learning systems, and later moved into product management. That journey taught me to work comfortably across research, engineering, design, and product, translating ambitious technical ideas into products people genuinely enjoy using.

I spent nearly five years at QuantumBlack, AI by McKinsey, helping Fortune 500 companies apply machine learning to complex problems across life sciences, financial services, and the public sector. Along the way, I evolved from forward-deployed engineering into product, leading AI developer tools that helped teams build, test, and scale machine learning applications more effectively.

In 2023, I joined PhysicsX as the company’s first product hire to help build an AI platform for mechanical engineering. Over the next two and a half years, I worked from the earliest ML research and product concepts through to a platform used by engineering teams across automotive, aerospace, and advanced manufacturing. As Head of Product, I built and scaled the product organisation, hiring across product and engineering as the team grew to more than 35 people, while helping the company grow through its Series A and Series B.

I later joined Tessl to build developer tools for AI coding agents, focusing on the workflows that help engineering teams understand, evaluate, and improve how agents perform across large codebases.

More recently, I started building Quark, an AI-native CAD product. Through it, I’m experimenting with how new software primitives, thoughtful interaction design, and AI can make professional mechanical engineering tools radically simpler.

I’m driven by curiosity and the pursuit of personal growth. I spend much of my free time reading and, more recently, learning to draw — partly for the joy of it, and partly as a reminder that great products begin with careful observation. Before you can simplify something, you first have to understand it.

.experience
Aug 2025 — Feb 2026

Product

Tessl

Built DevTools that made it easier for engineering teams to understand, evaluate, and improve how AI coding agents perform across large codebases.

Jan 2023 — Jun 2025

Head of Product

PhysicsX

Led product strategy and execution for an AI platform that helped mechanical engineers design better cars, aircraft, and industrial products.

Jul 2022 — Dec 2022

Product Manager

QuantumBlack, AI by McKinsey

Built AI developer tools that made it faster and easier for teams to develop machine learning applications.

Apr 2018 — May 2022

Lead Data Engineer

QuantumBlack, AI by McKinsey

Delivered machine learning solutions and data infrastructure as a forward-deployed engineer for Fortune 500 customers across life sciences and finance.

Nov 2017 — Mar 2018

Data Engineer

Keboola

Built data pipelines and analytics dashboards as a forward-deployed engineer using Keboola's platform.

.building

Products born from curiosity I couldn't ignore.

Quark generating a 3D keychain model from a text prompt

Quark Shipping

Building an AI-native CAD product that helps makers turn ideas into manufacturable designs in minutes. Through Quark, I'm exploring the software primitives, interactions, and design principles needed to make complex mechanical engineering tools radically simpler.

  • Multimodal AI
  • UX
  • CAD