Clay

Camp Clay awards szn 🏕️🥹🏆 For context: these awards come right after a Q2 where the team beat target by almost 70%, 71% of sellers hit quota, and ACV was up about 30%. So we got everyone in a room and yelled about the people who helped make it happen. Big huge congrats to: 🏆 Solange Levy - Top Quota Attainment Award 🌟 Marilin Berg - SL Rookie of the Half 🚀 Jake Rainess - SA Rookie of the Half 🧪 David Madding - Innovation & Exploration Award 🤝 Tyler Swanson - Partnerships Impact Award 🧱 Nate Segal - PipeGen Award 🛠️ Kyle Nicholson - Solutions Engineering Impact Award 📈 Brock Lownes - ClayDR Breakout Award 🔗 Michaela Pandorf - Product Partnership Award 👑 Sabrina Glaser - GTME MVP Award 🫶 Melina Giakas - Camp Counselor of the Year 🎤 Billy Parker - Top Elevator Pitch Couldn't be prouder of this crew 💚

I'm the Head of GTM Engineering at Clay. Here's every tool in our GTM tech stack for 2026: what we're building, what we're buying, and how we put it all together. We’ve tried to be intentional about buying best in class tools but building where we can get a unique advantage by doing so. I’ve marked ⭐ on pieces of our stack that are either Clay or internally built. We follow the “build on Clay” principle where we use Clay as infrastructure for many parts of our stack, but rely on end solutions for displaying data or providing interfaces. Our two biggest investments are Terra (our rep command center) and Quartz (our pricing agent). Building these tools aligns with our principles because how our reps work and how we price are two highly unique areas where the internal context we have beats what a vendor could do. But as you can see, the majority of our stack is bought - proof of the narrative violation that “we’re going to use AI to build everything internally”. Marketing - Inflection - lifecycle / campaigns / notifications ⭐ Clay Ads - ad audiences - LinkedIn / Meta / Google - ad execution - Webflow - Ploy - website conversion Prospecting - LinkedIn Sales Nav - manual prospecting - Exportly - Clay workflows in browser - Gong Engage - sequencing / ops triggered sequences ⭐ Clay Sequencer - rep triggered sequences ⭐ Terra / Terra MCP (internal rep UI) - signals, 1P data, call prep, ROI calculator Data - Snowflake - Fivetran - Segment ⭐ Clay Audiences Enrichment ⭐ Clay AI - Claude / ChatGPT / Slack - chat - Wispr Flow - voice - Codex / Claude Code / Cowork - automation and vibe coding - Notion - knowledge / context - Town - personal agents - Dust - team wide / org wide agents ⭐ Account Agents Orchestration ⭐ Clay Call recording - Granola - personal meeting capture - Gong - systematized meeting capture Sales - Salesforce - Claude Design - deck creation - Sigma - reporting and metrics CPQ - CaptivateIQ - commission and quota management ⭐ Quartz (internal agent) - pricing and scoping Got other tools we should try? Other ideas on where we can build? Let me know!

💬 Tip: Use Skills in Claygent Builder the next time you configure and test your agents in Clay. Claygent Builder is where your prompt, model, tools, and tests live in one place. Sooooooooooo, think of Skills as 𝘳𝘦𝘶𝘴𝘢𝘣𝘭𝘦 𝘮𝘶𝘴𝘤𝘭𝘦 𝘮𝘦𝘮𝘰𝘳𝘺 for your team’s know-how, e.g., a research sequence, tone check, or quality bar. This means: > You can pick a Skill mid-task > One agent can use multiple Skills > Build once, reuse across your team > Every run follows the correct set of rules Now available in every Clay workspace. ✌️

On behalf of ©2026 Clay Labs Inc., we would like to issue a formal apology. Earlier today, statements were made that did not adequately reflect the severity of the Workflows launch situation. This was an unacceptable oversight. As such, we are taking immediate corrective action, as per our CEO Kareem. Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Code Nodes Furthermore, we failed to properly communicate that the Clay CLI/API Agent Plugin can build, edit, and trigger Workflows. Corrective action: CLI/API Agent Plugin can build Workflows CLI/API Agent Plugin can edit Workflows CLI/API Agent Plugin can trigger Workflows CLI/API Agent Plugin can build Workflows CLI/API Agent Plugin can edit Workflows CLI/API Agent Plugin can trigger Workflows CLI/API Agent Plugin can build Workflows CLI/API Agent Plugin can edit Workflows CLI/API Agent Plugin can trigger Workflows CLI/API Agent Plugin can build Workflows CLI/API Agent Plugin can edit Workflows CLI/API Agent Plugin can trigger Workflows Finally, our prior messaging did not sufficiently emphasize observability at scale. Corrective action: Observability for every run Observability for every record Observability for every routing decision Observability for every agent decision Observability for every failure Observability for every credit spent Observability and scale Observability and scale Observability and scale Observability and scale Observability and scale Observability and scale We regret the error. While we have the floor, please direct your attention to clay.com/workflows for a very fun and interactive website experience. 🥺

𝗡𝗘𝗪: Introducing Workflows. 🎉🎉🎉 The fastest way to build GTM workflows in Clay. Workflows is Clay’s new orchestration layer: a node-based canvas where the data in Audiences turns into action. Wire triggers, conditional logic, agents, and your living GTM data layer together, then test, edit, and trace every run in one place. That gives GTM builders a richer, more editable way to: > 𝗧𝗿𝗮𝗰𝗲 𝗲𝘃𝗲𝗿𝘆 𝗿𝘂𝗻. See what came in, where it went, why it went there, what failed, how long it took, and what it cost. > 𝗧𝘂𝗻𝗲 𝘁𝗵𝗲 𝗹𝗼𝗴𝗶𝗰. Use deterministic rules, code, and Functions when the play needs precision. Bring in Claygents when it needs research, scoring, or classification. > 𝗕𝘂𝗶𝗹𝗱 𝗰𝗼𝗺𝗽𝗹𝗲𝘅 𝗚𝗧𝗠 𝗳𝗹𝗼𝘄𝘀 𝗳𝗮𝘀𝘁𝗲𝗿. Create Workflows from the CLI, chat with Sculptor, or convert an existing Table to get started. Then run Workflows with unlimited steps across unlimited records. (Tables live. Table limits die. ✌️) We’ve already seen the best GTM teams in the world use Workflows to hyper-simplify their routing-heavy plays and understand exactly why each record went where it went. 🚨 Our CEO, Kareem Amin, is going 𝗟𝗜𝗩𝗘 tomorrow, August 12 to talk about why and how we built Workflows. Then, we're going to demo real, production-grade builds!! Save your seat here → clay.link/Gxcpjv3 🦠

Okay who cares that Clay crushed Q2 and had our best quarter ever. It's Q3 now and we're back at 0.

there's only 1 reason u FREAKS would want that seat... the view is going to be peak. get your SCULPT tix at sculpt.clay.com

Celebrities are finally catching on to what I’ve been saying for 3 years

OpenAI recently posted about automating routine data reports (https://lnkd.in/eWNuf5P4). This is absolutely the way. We’ve been doing something similar at Clay (shoutout to Tessa Xie for leading this initiative!) on top of our agentic analytics infra. For anything we ship (product, campaign, feature, etc.), we can set up scheduled agent runs that look at relevant metrics, point out what moved (& hypothesize why), and send the people involved a report. You can also dig deeper by tagging in @monty. This saves each data scientist ~3–4 hours a week, but talking about time savings from AI is boring at this point. The more interesting impact is behavioral, because recurring work inevitably competes with fires and shifting priorities. A report meant for Monday can easily slip to Wednesday, and by Friday you start wondering if it’s even worth doing that week. That delay costs you a learning cycle. You notice + investigate something later, and make a change later. When you’re growing exponentially, moving that loop forward by a week can compound dramatically over time. Schedules guarantee cadence.