Home
Companies
Agentic Labs

Agentic Labs

AI system design tools for dev teams

Agentic Labs is making software design and architecture less painful by building AI into the process. Today, developers have no shortage of tools for instantly creating chunks of code, but designing the systems that integrate this code is the hardest and most manual part of software engineering. Our first product is Glide, an AI-first editor purpose-built for writing technical design docs. It connects to your codebase and saves your engineering team time on triage, planning, and implementation of complex code changes.

Agentic Labs
Founded:2023
Team Size:2
Location:New York
Group Partner:Tom Blomfield

Active Founders

Maksim Ivanov

Building self-assembling software. Previously at hedge funds and doing ML research

Maksim Ivanov
Maksim Ivanov
Agentic Labs

Robert McKnight

Worked on self-driving cars at Tesla and self-driving tractors at John Deere, with a detour to Bain for consulting after business school. Electrical Engineering and Computer Science (EECS) and Mechanical Engineering degrees from UC Berkeley and an MBA from Harvard Business School.

Robert McKnight
Robert McKnight
Agentic Labs

Company Launches

At Agentic Labs, we believe developers shouldn’t have to choose between thoughtfully designing code and shipping quickly.

Problem

Creating a good technical design makes the implementation 10x easier for most coding tasks, but it’s a slow and manual process. We’ve experienced this at companies like Bridgewater and SpaceX, where reliable and maintainable software is mission-critical.

Conversely, in a startup, you build super fast, but it’s easy to find yourself being crushed by the complexity you’ve created because you didn’t have time to write a spec or analyze 3 different approaches.

Until now, you’ve probably had to pick one of these two extremes.

Solution

We’re building Glide, an AI-first editor purpose-built for writing technical design docs. It integrates with your codebase, saving engineers time on triage, planning, and implementation of complex code changes.

Here’s how you use it:

1. Get started by adding a repo, setting the goal, and attaching code

We’re demonstrating on an open ticket in Vercel’s AI-Chatbot repo

2. Triage the task in one click

Jump-start your thinking. Refine and edit in-line.

https://youtu.be/2wpdJlTNkE8

3. Triage → Plan

From your triage, generate a step-by-step plan.

4. Plan → Implement

Once you like the plan, generate detailed code edits.

5. Send to VSCode to complete the change

Yes, we know, automatic PRs are coming soon.

https://youtu.be/0xz3uiHxg0c

Chat with a fully-informed AI assistant at any time

https://youtu.be/zSjvqpnuocI

Our ask