Clay Sculpt 2026: Workflows, Inbound SDR Agent, $7.1B Bet
On October 8, Clay stood in front of 2,500 people from 30 countries at Sculpt, its GTM conference in San Francisco, and announced that the spreadsheet is no longer the product. An agent-first interface where you describe outcomes instead of building tables. A visual Workflows canvas for orchestrating plays across millions of records. And its first agentic seller — an Inbound SDR Agent that qualifies leads by voice and demos your product in a sandbox.
The timing wasn't subtle. Weeks earlier, Clay closed a $115 million Series D led by Wellington Management at a $7.1 billion post-money valuation — more than double its $3.1 billion Series C from thirteen months before, on roughly $200 million in revenue and 17,000 customers. Sculpt was the pitch for what that multiple is buying: Clay as a self-learning revenue engine, not an enrichment tool.
As with every agent launch this fall, the announcements and the shipping dates tell different stories. Here's what's real today, what's behind a waitlist, and what the credit meter means for the bill.
The Three Launches
1. Agent-First Clay
The headline shift: instead of building a table and stacking enrichment columns, you give Clay's agent a goal in natural language — "find me ICP contacts at high-intent health tech companies and build a campaign to prospect into them" — and it plans the work itself. It sources companies, enriches contacts, builds audience segments, and drafts sequences.
Two mechanics matter here:
- Knowledge Hub. The agent reasons against a stored context layer — your ICP, buyer personas, positioning, and competitive intel. Campaign results feed back into it: Clay says the agent learns from outcomes and recommends Knowledge Hub updates, closing the loop between what you believe about your ICP and what actually converts.
- Goal-based check-ins. Demos showed the agent accepting a target like "50 more opportunities," building an execution plan, and checking progress on a cadence of every few days — the always-on pattern OpenAI shipped as Dots and Salesforce is building into Agentforce.
Availability: live for Sculpt attendees, rolling out to Clay's early access program "in the coming weeks." Everyone else waits.
2. Clay Workflows
A visual orchestration canvas that sits above the familiar table. Workflows support branching, looping, code execution, and agent-call steps across millions of records — the kind of multi-path GTM plays that previously meant chaining tables together and praying the credit math held.
Notable details:
- You can describe a play in natural language to Clay — or to the coding agent of your choice via CLI, a nod to the GTM-engineer crowd that already drives Clay from Claude Code and Cursor.
- Results write back into Audiences, so a workflow's output becomes targeting context for the next one.
- Account Agent steps handle in-depth account research mid-flow.
- Visual tracing lets you debug runs and bulk re-run failed branches.
Availability: beta now for all Clay users, with full launch slated for fall 2026. This is the one announcement you can actually touch today.
3. Inbound SDR Agent
Clay's first agentic seller, and the most ambitious reveal. The agent engages inbound leads in a voice conversation, qualifies them, answers product questions, and — the genuinely new part — demos your actual product in a sandboxed environment, with a use case tailored to what it learned about the lead. Qualified buyers get routed to a rep.
How it learns: from your existing call recordings of top sellers, plus demo videos and documentation. In the Sculpt demo, the agent spotted an open accounting role at a prospect's company and built a customized sandbox demo around it — capturing a lead that generic scoring would have discarded. Clay is pitching it at hybrid PLG/sales-led motions where inbound volume outstrips rep capacity, and claims the agent's conversations feed recommendations back on ICP, positioning, and competitive differentiation.
Availability: pre-beta partner waitlist, shipping in phases starting fall 2026. This is a preview, not a product. For comparison, Salesforce's inbound equivalent — Piper, the ex-Qualified agent — is GA today with chat-based qualification; nobody has shipped voice-plus-sandbox-demo at GA yet.
The Same 48-Hour Wave, Again
Sculpt didn't happen in a vacuum. The week of October 6 was the densest agent-launch window of the fall:
| Vendor | What shipped | Status |
|---|---|---|
| Clay (Oct 8) | Agent-first interface, Workflows, Inbound SDR Agent | Early access / beta / waitlist |
| Apollo (Oct 8) | AI GTM System with Intelligence Layer | Live; Builder Studio later |
| ZoomInfo (Oct 1) | Agent Teams, after acquiring DoubleO.ai Sept 30 | Live, no new contract required |
| Gong | Enrich and Deep Mode GA | Live, metered on Gong Credits |
We covered the first round of this arms race in Gong, Apollo & ZoomInfo All Shipped AI Agents in 48 Hours. The pattern holds: every GTM data vendor is racing to become the reasoning layer, and every launch pairs a bold agent demo with a credit meter.
It's worth keeping one skeptical observation in view. Industry reviewers who sat through dozens of these vendor pitches this fall note that nearly every one now demos some form of context graph or knowledge layer — and almost none have shown measurable outcome differences against plain retrieval. Knowledge Hubs are the new pitch; lift is still mostly asserted, not published.
What This Costs (Nobody Said)
Clay published no pricing for any of the three launches. The agent-first interface runs on credits — same as everything in Clay — with no published rates for agent actions yet.
That matters because credit burn is already the core tension in Clay's model. Our Clay pricing breakdown walks through how waterfall enrichment routinely lands teams at 3x the headline plan price, and our Clay review documents the $500+/month burn patterns on moderate usage. An autonomous agent that sources, enriches, and sequences on its own initiative is a credit meter with its hands on the wheel. Until Clay publishes agent-action pricing, assume the goal "50 more opportunities" has an unbounded denominator.
The Series D math frames the incentive: a 35.5x revenue multiple is priced for aggressive expansion, and usage-based agents are the cleanest expansion lever in SaaS right now. Gong made the same move with Gong Credits; Microsoft meters its Sales Development Agent per worked lead. The agent era's business model is: the demo is free, the autonomy is metered.
Should Your Team Care?
If you're a Clay customer today: turn on the Workflows beta. It's the shippable piece, it addresses a real pain (multi-table play sprawl), and the visual tracing alone will save your GTM engineer debugging time. Treat the agent-first interface as a roadmap item and watch your credit dashboard when you get access.
If you're evaluating Clay against alternatives: nothing announced at Sculpt changes the near-term comparison, because the differentiating pieces are waitlisted into fall. The Clay alternatives landscape and the MarketBetter vs Clay comparison still come down to the same question: do you want a flexible enrichment workbench you assemble (and meter) yourself, or an opinionated system that runs the workflow end to end?
If the Inbound SDR Agent is the draw: join the waitlist, but benchmark it against what's GA. Voice qualification plus sandboxed demos is a genuinely new combination — and also two hard problems (voice latency and demo-environment provisioning) stacked on top of each other, shipping "in phases."
The strategic read: Clay's "self-learning revenue engine" — audiences, agents, sequencing, and ads in one loop that learns from campaign results — is a vision we'd endorse, because it's the architecture MarketBetter already runs. Signal intelligence identifies in-market accounts, visitor identification catches them on your site, and Loops handles the nurture follow-through — sending only when a real reason exists, not on a calendar. The difference is packaging: one system at $99/user/month, published, instead of a credit meter attached to a waitlist. If you'd rather run the loop today than join a pre-beta for it, book a demo.
FAQ
What did Clay announce at Sculpt 2026? Three things: an agent-first interface where you describe goals in natural language and Clay's agent plans and executes the work against a Knowledge Hub; Clay Workflows, a visual canvas for branching, looping, multi-agent GTM plays (beta now, launching fall 2026); and an Inbound SDR Agent that qualifies inbound leads by voice and demos the product in a tailored sandbox (waitlist, shipping in phases from fall 2026).
What is Clay's Inbound SDR Agent? Clay's first agentic seller. It holds a voice conversation with inbound leads, qualifies them, answers questions, and demonstrates your product in a sandboxed environment customized to each lead's use case, learning from your top sellers' call recordings. Qualified buyers are routed to reps. It's currently pre-beta behind a partner waitlist.
Is Clay Workflows available now? Yes, in beta for all Clay users as of Sculpt (October 8, 2026), with full launch planned for fall 2026. It supports branching, looping, code execution, and agent steps across millions of records, plus natural-language play creation via Clay or an external coding agent over CLI.
What is Clay's valuation in 2026? $7.1 billion post-money, after a $115 million Series D led by Wellington Management in September 2026 — up from $3.1 billion thirteen months earlier. Reported revenue is roughly $200 million across 17,000+ customers.
How much do Clay's new agents cost? Clay hasn't published pricing for the agent-first interface, Workflows, or the Inbound SDR Agent. Agent actions run on Clay's existing credit system, where real-world costs routinely exceed headline plan prices — see our Clay pricing breakdown for how the credit math works.