★ INSERT COIN◆NOW PLAYING: VENTURES◆HIGH SCORE: $100M ARR◆★ NEW STAGE UNLOCKED: ABOUT ME◆PRESS START◆★ DEMO DAY 04:00:00◆
★ INSERT COIN◆NOW PLAYING: VENTURES◆HIGH SCORE: $100M ARR◆★ NEW STAGE UNLOCKED: ABOUT ME◆PRESS START◆★ DEMO DAY 04:00:00◆
◀ BACK TO FEED
NEWS★ DEVELOPER TOOLSOCT 26, 2026

Paradev Raises $11.5M for Coding Agents

Paradev, formerly Superset, raises an $11.5 million seed round led by Union Square Ventures to orchestrate multiple AI coding agents.

Paradev Raises $11.5M for Coding Agents

Developers are increasingly managing several AI coding agents at once, but coordinating their tasks and checking their output can become a workflow of its own. Paradev is building an environment for that shift.

What happened

Paradev, previously called Superset, announced an $11.5 million seed round led by Union Square Ventures. Backers also include Y Combinator, BoxGroup, 468 Capital and other investors and founders.

The startup offers a workspace where engineers can run multiple coding agents, including Claude Code and Codex, on separate tasks in parallel. It helps manage workspaces, branches, agent activity and review before code changes are integrated.

Paradev says hundreds of thousands of developers in 180 countries have used its product, including users at about one in eight Fortune 500 companies. Those are company-reported adoption indicators, not confirmed enterprise-wide customer contracts. The announcement also marks a rebrand: existing Superset accounts and workspaces carry over to Paradev.

Why it matters

Running parallel agents can accelerate coding work, but raises coordination problems around duplicated effort, conflicting edits, test coverage and human review. A management layer that works across different models could make these workflows less dependent on any single AI provider.

The risk is that model companies and integrated development environments may incorporate similar functions directly, limiting space for an independent orchestration platform.

The bigger picture

The coding-tool market is moving from AI assistance inside an editor toward supervised teams of software agents. If that transition continues, tooling for assigning work, maintaining context and validating results could become as important as the models writing code.