AI Agents Drive DevTools Consolidation Wave

The devtools world didn’t slow down during our two-week break. While we were away, AI agents and developer infrastructure grabbed headlines with over $2.5 billion in fresh funding pouring into the sector. This edition covers everything from enterprise AI platforms hitting production-ready status to real-time analytics tools that are powering the next wave of applications.

What we’re seeing is devtools consolidation changing how teams build, deploy, and scale intelligent systems.

🏆 Funding Wins

Mistral Marks €1.7B Series C To Multiply AI Momentum

Mistral AI has closed a €1.7 billion Series C at an €11.7 billion valuation, with semiconductor equipment leader ASML Holding taking the lead. The strategic partnership signals Mistral’s pivot toward custom enterprise AI solutions, particularly targeting complex engineering challenges in the semiconductor value chain. Founded by Arthur Mensch, Timothee Lacroix and Guillaume Lample, Mistral has positioned itself as Europe’s frontier research alternative with its open-source approach. The ASML partnership suggests Mistral’s models will tackle specialized use cases like process optimization rather than general-purpose applications. The round included participation from NVIDIA, Andreessen Horowitz, DST Global, General Catalyst, Lightspeed, Index Ventures and Bpifrance.

🤖 Read Mistral AI’s announcement.

Recall.ai Raises $38M Series B for Conversation Data Infrastructure

Recall.ai, founded by David Gu and Amanda Zhu, has raised a $38 million Series B at a $250 million valuation. Bessemer led the round with participation from HubSpot Ventures and Salesforce Ventures. Recall.ai provides APIs that developers use to record and process conversations across Zoom, Google Meet, dialers, and even in-person meetings. The company already supports over 2,000 clients and processes millions of calls monthly, abstracting away a huge technical lift.

“99% of the context AI needs is never written down — it’s spoken.” – Gu

With this funding, Recall is expanding into new real-time SDKs and storage to power the next generation of conversation-native apps.

🗣️ Congrats to the Recall team, learn more.

Alpic Ascends with $6M To Advance Agent-First Cloud

French-founded Alpic has emerged with a $6 million pre-seed, backed by Partech, K5 Global, and angels from Mistral, Datadog, and Dataiku. Founded by Charles Sonigo, Pierre-Louis Theron, Nikolay Rodionov, Erica Beavers and Frédéric Barthelet, Alpic is building the first MCP-native cloud platform designed specifically for agent infrastructure rather than retrofitting existing systems. The team previously built video delivery startup Streamroot, helping media companies transition from broadcast to streaming. Now they’re applying similar infrastructure thinking to the agent shift. Alpic’s platform lets developers deploy and manage MCP servers in minutes with built-in security, analytics, and production tooling.

🏔️ Congrats to the Alpic founders. Try the public beta here.

Fiveonefour Fetches $17M To Fuel Fast AI Analytics

fiveonefour has secured $17 million in funding led by Dimension Capital with support from Stage 2, Flybridge, Ridge, Tokyo Black and Vermillion Cliffs. The Portland-based startup, founded by Tim Delisle (former VP of Engineering at Nike), is tackling the data infrastructure demands of AI and real-time analytics. Its open-source Moose Stack, built in TypeScript and Python, handles data ingestion, transformation, and real-time querying so developers can quickly embed AI-ready analytics into their apps without managing complex data pipelines. Customers like F45 Training use the platform to drive personalized workout insights, reporting members who engage with Fiveonefour-powered features are 2.5x more valuable. The funding will accelerate enterprise adoption and grow its Boreal cloud and Sloan AI agents.

📊 Kudos to Fiveonefour, give the Moose Stack a look.

Runware Races Ahead With $13M To Reinvent AI Media

Runware, founded by Ioana Hreninciuc and Flaviu R. in 2023, has just closed a $13 million Seed round led by Insight Partners to grow its developer-first AI media platform. The company already powers more than 400,000 models, serving 100,000 developers who reach 250 million end users. Its Sonic Inference Engine lets teams switch models instantly, unify costs, and chain image, video, and text generation without juggling multiple APIs. Platforms like Quora, NightCafe, and OpenArt rely on Runware for speed and stability. The fresh capital will support research in audio generation and LLM integrations. By doubling down on transparent pricing and fast inference, Runware is positioning itself as the go-to API for media-first AI apps.

🏃🏾 Hats off to Runware — check the news.

Cognition AI Scores a Staggering $400M To Cement $10.2B Valuation

Cognition has raised over $400 million in Series B funding led by Founders Fund with participation from Lux Capital, 8VC, Elad Gil, Definition Capital and Swish Ventures. Just two years old, the San Francisco startup has quickly emerged as a leader in AI coding agents. Its flagship tool Devin, “the world’s first AI software engineer,” has driven annual recurring revenue from $1 million in late 2024 to $73 million this summer. With fresh backing, Cognition will expand engineering and accelerate product development around Devin and its Windsurf acquisition.

💻 Congrats to Scott Wu and team. Read more here.

Series D and UP ⬆️

The past fortnight has also brought two blockbuster late-stage raises.

Anthropic closed a $13 billion Series F at a staggering $183 billion valuation, co-led by Iconiq and Fidelity. The company says funds will deepen safety research and support global enterprise demand as Claude adoption surges past 300,000 businesses.

Replit landed $250 million at a $3 billion valuation. Replit has reported annualized revenue growth from $2.8 million to $150 million in under a year and unveiled Agent 3 last week; its latest AI ‘collaborator’ for testing, debugging and automation – see more below.

😲 Catch more from Anthropic | Explore Replit

🚀 3, 2, 1… Launches

Nottelabs Launches Web Agents For Enterprise

Nottelabs (YC S25), founded by Andrea P. and Lucas Giordano , has launched a full-stack platform for web agents that automates repetitive workflows without breaking. Their “Agent Workflows” lets developers freeze repeatable steps and keep agents only where they add value, creating automation that adapts when websites change. The platform includes stealth browser sessions, secure vaults for MFA and SSO, personas for account creation, and even payment handling.

🌐 Sign up and try Agent Workflows. The team is calling on developers tackling complex web tasks to test and collaborate.

Atla Aids AI Agent Reliability

Atla, co-founded by Roman Engeler and a small team of AI researchers, has launched a platform to help teams diagnose and fix recurring failures in their AI agents. Instead of showing endless traces, Atla automatically flags root causes, clusters issues into patterns, and produces actionable fixes. Startups using the platform report spotting failures weeks faster and doubling the number of shipped improvements. Drawing on Roman’s research into iterative model self-improvement at Stanford, Atla is focused on reliability in high-stakes agent systems. Teams can pinpoint the “2–3 failures that actually move the needle,” speeding deployment cycles and improving trust in their agents.

🔎 Find your agent’s weak spots with Atla – give them a try.

Xpander.ai Launches Full‑Stack Platform for AI Engineers

Xpander.ai, founded by David Twizer and Ran Sheinberg, has gone live on Product Hunt where it picked up Product of the Day, Week and Month. They’ve told us to think of it as the “Supabase of AI agents”, the platform gives engineers everything needed to take agents from demo to production in hours. It ships with MCP connectors and registry, a stateful DB, AI gateway, rollback‑ready test environments, runtime and MCP hosting. Developers get a visual workbench that never locks them out of code, a large pre‑built tools library, and fully self‑service deployment options via xpander Cloud or their own Kubernetes cluster.

⚙️ Give xpander.ai a go and show your support on Product Hunt

Replit Unveils Agent 3 For Automated Testing and Debugging

Replit has launched Agent 3, its latest AI collaborator that automatically checks buttons, forms, APIs, and data sources in live browsers while spotting errors and applying fixes without manual intervention. The agent transforms development workflows by handling testing and debugging autonomously.

Key capabilities:

• Runs autonomously for 200+ minutes

• Self-testing mode in live browsers

• 10x more autonomous than previous versions

• Builds automations from Slackbots to automated reports

🤖 Meet Agent 3 and see what automated development looks like.

🤝 Deals and Acquisitions

OpenAI Acquires Statsig, Reshuffles Leadership

OpenAI is acquiring product testing startup Statsig for $1.1 billion in an all-stock deal. Statsig founder Vijaye Raji will become CTO of Applications, reporting to Fidji Simo. The move aims to accelerate product development for ChatGPT and other AI apps. Leadership changes include Kevin Weil leading OpenAI for Science and Srinivas Narayanan heading B2B applications.

🧪 Read Statsig’s founder post.

CoreWeave Buys OpenPipe For Agent Training

CoreWeave has acquired OpenPipe, a YC-backed startup specializing in reinforcement learning tools for AI agents. OpenPipe’s open-source ART toolkit helps enterprises develop customized agents, aligning with CoreWeave’s high-performance AI cloud. This acquisition expands CoreWeave’s offerings, following its earlier purchase of Weights & Biases.

🪈 More on the OpenPipe deal.

Kong Acquires OpenMeter For API Monetization

Kong Inc. has acquired OpenMeter, an open-source platform for usage-based metering and billing. The integration into Kong Konnect will enable organizations to monetize APIs, LLMs, and data streams at scale. OpenMeter’s team, including co-founders Peter Marton and Andras Toth, will join Kong.

🦍 See Kong’s announcement.

🔦 DevTool of the Week: InKeep

AI agents are everywhere, but most tools split teams apart. Developer-focused platforms shut out business owners, while no-code builders leave engineering teams on the sidelines. Inkeep is our DevTool of the Week because it tackles this exact problem, making collaboration possible across the stack.

Founded in 2022 by Nick G. (CEO) and Robert Tran (CTO), InKeep now has 15 employees based in San Franciso. Fresh off a $13 million seed led by Khosla Ventures and GreatPoint, the company has now launched Inkeep Agents. The platform lets engineering and non-technical teams build AI agents side by side, without compromise. Agents can be created visually in plain English, or in code using a first-class SDK. Either way, they stay editable both ways, meaning teams can switch seamlessly between visual and technical modes.

For developers, Inkeep ships with RAG, multi-agent architecture, MCP servers, chat UIs, and an open source SDK ready to integrate. For non-technical users, the Visual Builder opens up customer-facing automation. Real-world examples include support bots that process refunds, agents that turn tickets into docs, and onboarding copilots for product adoption. Customers like Anthropic, Midjourney, PostHog, Clay and Solana are already putting these to work.

With built-in auditability, testing, and transparent “don’t know” responses, Inkeep prioritizes trust while enabling speed. It is a platform where every team can actually collaborate on agents that get things done.

👏👏 Explore Inkeep

Got a launch, raise, or tool we should feature next week?

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