Over $1Bn Funding in AI Agents, QA Infrastructure and DevOps

Over $1 billion was raised by devtools companies week, with the biggest checks going to AI agents, QA infrastructure, and DevOps automation. Databricks closed a massive round to fund agent platforms, while companies like Functionize, SRE.ai, and Phoebe raised millions to build self-healing testing, AI-powered DevOps workflows, and software monitoring. Meanwhile, launches focused on practical wins: Parag Agrawal’s Parallel Web Systems rethinking web agents, ZeroEval shipping self-improving agent evaluations, Momentic capping launch week with mobile testing that delivers near-instant emulators, and yes, even “vibe tools” like Vibe Code Go that let developers code from their phones. – It’s all here.

🏆 Funding Wins

SRE.ai Raises $7.2M To Put AI Agents Into Enterprise DevOps

SRE.ai, founded in 2024 by former Google Research and DeepMind engineers Edward Aryee and Rajsekhar Kadiyala, announced a $7.2 million seed round led by Salesforce Ventures and Crane Venture Partners. The company builds AI-native DevOps agents that connect with tools across AWS, Salesforce, and ServiceNow to handle workflows like testing, releases, monitoring, and security checks through a chat-like interface. Aryee said the idea grew gradually:

“It wasn’t one big lightbulb; it was death by a thousand cuts.”

👏 Congrats SRE.ai — learn more here.

Databricks in Process of Closing $1B At $100B Valuation For AI Agent Push

Databricks, founded in 2013 and best known for its data analytics solutions, is closing a MEGA near-$1 billion round at a $100 billion valuation, sources told TechCrunch and the Wall Street Journal. The primary raise was co-led by Thrive and early backer Insight Partners. The company has now raised about $20 billion to date and held employee secondaries earlier in 2025. Databricks says the fresh capital will fund two projects it launched in June: Lakebase, an enterprise Postgres-based database for AI agents, and Agent Bricks, an AI agent platform.

They also announced last week plans to acquire machine learning startup Tecton “to expand its AI agent offerings, the latest in a string of deals aimed at offering full-scale AI building tools for enterprise customers.” (Reuters)

🧱 Congrats Databricks — read the WSJ report here and the TechCrunch article here.

Firecrawl Raises $14.5M To Power Developer And Agent Crawling

Firecrawl, founded in 2022 and based in San Francisco, builds an open source web crawler and a commercial API used by developers, AI agents, and companies like Shopify, Replit, and Zapier. The startup announced a $14.5 million Series A led by Nexus Venture Partners, with participation from Shopify CEO Tobias Lütke and Y Combinator. Firecrawl says 350,000 developers use its tools and it is already profitable. The team is rolling out search and natural language prompt support in its API and exploring ways to pay publishers when AI reuses their content. Co-founder and CEO Caleb Peffer joked about pitching Nexus:

“I actually fell out of my chair. And Abhishek, as a great investor does, caught the chair and me as I was falling.”

🔥 Nice win Firecrawl — read the announcement here.

Functionize Nets $41M To Expand AI Testing And QA

Functionize, a San Francisco company offering an AI platform for automated software testing, raised $41 million in a Series B led by Mumford Investments and LHH Investments, with participation from Canva and Wipro venture arms. Led by CEO, Tamas Cser, the platform generates and self-heals test scripts from natural language descriptions, creates test data, and runs API and performance checks across global VMs.

Functionize says customers ran over 1 billion AI-powered tests last year and include enterprises such as GE Healthcare and ServiceNow.

🧪 Congrats Functionize — learn more here.

TensorZero Scores $7.3M To Open-Source LLM Engineering Stack

TensorZero, founded in January 2024 and building an open-source stack for production LLM applications, closed a $7.3 million seed round led by FirstMark with participation from Bessemer, Bedrock, DRW, Coalition, and others. They published their stack in September 2024 and says it unifies an LLM gateway, observability, optimization, evaluation, and experimentation tools so teams can build feedback loops from production signals. TensorZero plans to use the funding to accelerate development of its open-source components that help LLMs learn from real-world experience and to grow engineering in NYC.

0️⃣ Congrats to CEO Gabriel Bianconi, CTO Viraj Mehta and the rest of the team — read their post here.

Phoebe Raises $17M To Build A Software Immune System

Phoebe, a London startup founded in 2024 by Matt Henderson and James Summerfield, secured $17 million in a seed round co-led by GV and Cherry Ventures and launched its platform publicly. Phoebe uses swarms of AI agents to monitor logs, traces, commits, and metrics in a read-only, vendor-agnostic way. The agents diagnose incidents, generate code or configuration fixes, and in some cases predict problems before they cause downtime. Early customers include Trainline and PPRO. GV investor Roni Hiranand said:

“AI has transformed how code is written, but software reliability has not kept pace. Phoebe is building a missing layer of contextual intelligence.”

🎉 Congrats Phoebe — check their launch here.

Zed Zooms Ahead with $32M To Build Collaborative AI Native Editor

Zed, the open-source, multiplayer code editor built from the ground up in Rust, raised $32 million in a Series B led by Sequoia Capital, bringing total funding to over $42 million. Based in Boulder, the project was started by creators of Atom and Electron and is led by CEO and co-founder Nathan Sobo. Zed focuses on persistent, contextual conversations tied to code using character-level permalinks and CRDTs, and it adds real-time collaboration for humans and AI agents. Since open-sourcing in 2024, Zed has drawn over 150,000 active developers and 1,100 contributors. “We’re building fundamentally better primitives for collaborative coding,” Sobo said.

👩🏽💻 Congrats Zed — read the blog here.

🚀 3, 2, 1… Launches

Parag Agrawal’s Parallel Web Systems Nets $30M To Rethink Web Agents

Parallel Web Systems, founded by former Twitter CEO Parag Agrawal, launched with $30 million to change how AI agents interact with the web. The company offers a cloud research platform running eight AI research engines and says its stack outperforms top models like GPT-5 on web research benchmarks. Parallel positions itself against browser infrastructure providers and large LLM-driven search agents from firms like OpenAI and Google. The launch frames Parallel as a web-native research and agent play aimed at faster, more capable web-scale agent workflows.

👀🤩 We’re loving what we see on their website.

ZeroEval Ships Self Improving Agent Evals

ZeroEval, launched by Sebastian Crossa and Jonathan Chávez (YC S25), helps teams evaluate and improve AI agents using human feedback. The tool builds calibrated LLM judges that learn from labeled mistakes and production data, plus Autotune for automatic evaluation and prompt optimization across many models. ZeroEval targets long-running, multi-turn agents and tasks with subjective outputs where static evals fail. The founders previously built llm-stats.com and bring experience from Micro and Datadog. The product aims to turn messy, high-friction eval cycles into a continuous learning loop so agents steadily match human preferences in production.

0️⃣ Nice launch ZeroEval — try it or book a demo.

Observee Provides Integration Engine For Vertical AI

Observee, led by Rishabh Chanana with cofounders Chris and Rajmeet, launched a unified integration platform for Vertical AI builders. The product pairs a Data Ingestion Engine that pulls data and builds agentic search from any platform with an Action Engine that executes tasks via managed APIs and robust browser automations. Observee combines computer vision and DOM parsing to keep automations reliable and uses semantic search plus a knowledge graph to serve agents real-time, contextual data. The startup targets AI for legal, finance, healthcare, and other verticals where integrations are the bottleneck to productising agents.

👁️ Well done Observee — give it a go here.

Vibe Code Go Lets You Code From Your Phone

Vibe Code Go, founded by Chris Nolet (YC S25), is a mobile-first app that lets engineers “vibe code” on iOS with a macOS companion. The app treats coding like conversational messaging, surfacing multiple chat threads and interruptions so you can dictate features, iterate with an agent, and push changes to your local repo. It runs Claude Code on a Mac under the hood and syncs edits straight to disk. Vibe Code Go aims to make small, meaningful dev work possible away from a desk, trading big-screen context for fast agentic workflows and immediate feedback.

👏 Try Vibe Code Go — download and give it a spin.

🔦 DevTool of the Week: Momentic

Momentic grabbed our attention with a packed launch week and daily releases that show a clear product rhythm and some great ‘momentum‘ 😉.

Founded by Wei-Wei Wu (CEO) and Jeff An (CTO) in late 2023, Momentic uses AI to let developers write end-to-end tests in plain English and keep them working as apps evolve. The team raised $3.7M in March and says customers see about 60% less debugging time and 30% faster test runs. That traction matters when you run tests across hundreds of thousands of screens and hundreds of millions of interactions a month.

What shipped during launch week?

• Day 1 Copilot & MCP server: IDE-first test creation and refactoring. AI models can observe pages, preview interactions, and generate or modularize tests.

• Day 2 Quarantine: One-click isolation for flaky tests, custom thresholds and Slack alerts to stop noisy tests from blocking merges.

• Day 3 Suggestions and GitHub app: Automatic PR context, execution insights, and one-click test edits to reduce drift.

• Day 4 Failure Recovery: Agentic auto-recovery and custom instructions so transient UI or network issues do not turn into real regressions.

• Day 5 Mobile: The BIG one! Plain-English mobile tests, blazing-fast emulators, native and WebView context switching, and near-instant interactions.

Why it matters? Momentic removes brittle selector work and constant context switching, so engineers stay in flow and shipping speed improves. The mobile launch in particular tackles slow, flaky, expensive testing with impressive latency numbers. Worth watching for teams that want reliable, agent-driven test pipelines.

⏩ We’re loving this Momentic. Read the blog and check out the momentum here.

 

Got a launch, raise, or tool we should feature in our next issue on the 15th September?

Reach out to one of our team.

Louise Ogilvy (Recruitment Director) – UK/USA

Becca Combe MIRP CertRP (Lead Recruiter) – USA

Ben Chipchase (Lead Recruiter) – UK

Natalie Harper (Marketing Administrator)

Looking for your next DevTools role?

Visit devtoolsjobs.com