This week’s funding centres on two problems that keep surfacing as AI moves into production: observability infrastructure that breaks under the volume and complexity of agent-generated data, and evaluation tooling that can actually tell whether agents are working correctly. Several new launches address the same gap from the build side, with new frameworks and execution layers aimed at teams running multiple agents in live environments. Worth noting on the side: Groq confirmed a $650 million raise this week! Here are the funding wins we saw in the devtools start up world:
🏆 DevTools Funding Wins
Tsuga Raises $35 Million to Deploy Observability Infrastructure Inside the Customer’s Own Cloud
Gabriel-James Safar and Sébastien Deprez co-founded Tsuga in Paris in 2024 to eliminate the core flaw in legacy observability: telemetry leaving the customer’s environment entirely. Singular led the $35 million Series A, with General Catalyst, DST Global Partners, QuantumLight, Picus Capital and Databricks Ventures participating. The platform deploys inside the customer’s own cloud across Azure, AWS, GCP and sovereign environments with no sampling, no duplication and a single per-GB pricing model. Six months out of stealth, Tsuga has reached several millions in ARR with average contract values in the six figures. 🔍 Congrats to the team. Read the full announcement here.
Patronus AI Closes $50 Million Series B for Simulated Agent Evaluation Environments
Founded in 2023 by former Meta AI researchers Anand Kannappan and Rebecca Qian, Patronus AI builds simulated digital environments where agents are stress-tested using reinforcement learning before reaching production. The platform surfaces shortcuts and failure modes that standard benchmarks miss, and virtually every frontier AI lab is now a customer. Revenue has grown 15-fold over the past year. Greenfield Partners led the $50 million Series B, with Notable Capital, Lightspeed, Datadog and Samsung participating. Total funding now stands at $70 million. 🧪 Well done to Anand, Rebecca and the team. Read the announcement here.
DevPlan Raises $2.5 Million to Automate the Coordination Work AI Coding Has Left Behind
Chris Bee and Anton Safonov co-founded Devplan in 2025 after running engineering teams at Uber, Amazon, Snap and Meta. Their product Weaver connects GitHub, Jira, Slack and meeting tools to generate daily digests, answer status queries and flag risks automatically, without anyone filing an update. Queries run through a pre-built knowledge graph, running roughly twice as fast and three times cheaper on token costs than scanning raw data on every request. AI2 Incubator led the $2.5 million seed, with Acequia Capital, Mighty Capital, Grand Ventures and eLab Ventures also contributing. 🗂️ Wonderful news DevPlan. Find out more about DevPlan here.
Sazabi Secures $8 Million Seed Round and Opens Beta for Its AI-Native Observability Platform
Sherwood Callaway, a second-time YC founder who previously built infrastructure and observability teams at Brex, closed an $8 million seed and opened Sazabi to public beta this week. J2 Ventures, Village Global and Y Combinator co-led the round, with 60+ angels including Harrison Chase of LangChain, Matt Biilmann of Netlify and Paul Klein of Browserbase. Sazabi monitors systems using logs alone and uses AI to detect anomalies, investigate and propose fixes. In under a month, 50 teams including Mintlify, Daytona and Mastra are already in production. 🔥 Huge congrats to Sherwood and the team. No more waitlist — book a demo here.
🚀 New DevTools Launched
Magnitude Debuts a Coding Agent Built Entirely on Open Models at 60% Lower Cost
Tom Greenwald and Anders Lie launched Magnitude, a coding agent running entirely on open models at pass-through API cost with no markup. The headline claim is 60% cheaper than Claude Code with no performance drop, benchmarked across 100 trials on two real projects. GLM 5.2 leads sessions, DeepSeek V4 Flash handles exploration, and Kimi K2.7 Code covers implementation and review. Inference runs on Fireworks AI and Wafer with zero data retention. 🚀 Great work Tom and Anders. Try it now
Linzumi Launches Team Chat for Directing Multiple Coding Agents at Once
Sean Grove, a three-time YC founder who previously built OneGraph (acquired by Netlify) and worked on post-training and alignment at OpenAI, launched Linzumi as the coordination layer missing from multi-agent workflows. Teams kick off work, review agent output and ship from a single chat interface. A partnership with Wafer.ai gives new accounts free daily access to GLM 5.2. Currently works with Codex, with Claude Code support coming. 👏 Congrats Sean. Try it free at linzumi.com.
AgentX Releases an Evaluation Framework for Testing AI Agents Before Production
AgentX AI launched an evaluation framework built around the idea of CI/CD for agent deployments. Teams can create test suites, compare model performance across providers, trace failures end-to-end and get AI-generated root cause analysis with suggested fixes before anything reaches production. Founder, Xuelai (Robin) Wang announced that it ranked number one on Product Hunt on June 22nd. 🔧 Well done to the team. Explore the platform and show your support on Product Hunt.
Skybridge Open Sources a React Framework for Building MCP Apps Across AI Assistants
Alpic launched Skybridge V1.0, a React framework for building MCP apps with developer experience at its core. V1.0 brings a cleaner type-safe API and dev tools rebuilt from the ground up, including server tunneling, testing and audit options for app developers and their agents. It abstracts compatibility differences across Claude, ChatGPT, VS Code and other MCP clients so teams can code once and ship everywhere. The framework already has 100,000 monthly downloads and is in use across Fortune 500s and early-stage startups. Julien Vallini, founding engineer at Alpic, took it to Product Hunt this week with a simple framing: “The internet is going headless. MCP Apps are the new UI. Every host has its own quirks, and writing the plumbing to support all of them is painful. So we built Skybridge.” 🌉 Congrats to the Alpic team. Support the launch on Product Hunt here.
📆 Upcoming Events
DevRelCon NYC 2026 | July 22–23, 2026 | Industry City, Brooklyn, NYC | Talks and workshops on AI developer tooling, agentic platforms, DevEx, and developer adoption strategies. 🎤
WeAreDevelopers World Congress 2026 | 8–10 July 2026 | Berlin, Germany | 500+ speakers across 20+ stages covering AI-era engineering, agentic systems, platform engineering, security, and developer tooling. 15,000+ builders expected. 🌍
⚒️ DevTools Job Opportunities of the Week
Senior Solutions Engineer | London – OTE £190k | Texas, Chicago, New York, Boston – OTE $220k
-
The Role Senior customer-facing pre-sales role for a developer-first AppSec platform.
-
What You’ll Do Lead enterprise sales cycles, run Proof of Value engagements, and integrate the product into CI/CD pipelines and developer workflows.
-
Requirements 6 to 8 years in customer-facing or pre-sales engineering, with strong depth across DevOps, AppSec, and CI/CD.
-
Best Fit Engineers who can read code, understand vulnerabilities, and speak credibly with senior technical stakeholders.
Contact Louise Ogilvy for more information and to apply.
Got launch news or funding to share? Send it to Natalie Harper.
Want more details on the opportunities here or want to register your resume with us, reach out to Becca Combe MIRP CertRP and Louise Ogilvy directly.
AI, observability, agents, devtools, funding, infrastructure, evaluation, coding, developers, security, tooling, production, platform
