When we opened the Rally Agent & MCP beta in August, we shared our bet: the future of research operations is agentic. Researchers and product builders shouldn't spend their days pulling lists, sending mail merges, coordinating calendars, and managing incentives just to talk with users. At Rally, we believe that work should simply be done for them, safely and within the guardrails an enterprise requires.
Six weeks later, that future isn't a bet. It's how our customers are actively working.
Today, Rally Agent and MCP are fully released and accessible to all customers.
Rally Agent and MCP are built to end manual recruitment - the slowest, most manual phase of user research, which typically spans ten steps and six apps per study. Now, researchers and product builders simply describe what they want to learn, and Rally builds the study from approved templates, recruits from the company's own panel, and automatically handles outreach, scheduling, and incentives. Every action runs inside the organization's existing governance controls.
"Trusting an agent with recruitment is a very high bar," said Oren Friedman, co-founder and CEO of Rally. "An agent has to contact real customers, handle sensitive data, manage consent, and issue incentives. We’ve spent five years building the infrastructure that makes that safe at scale, which is why Rally customers like Calendly, Figma, Ramp and others, trust the Rally Agent and MCP to run high quality research programs."
During the beta, the research team behind one of Included Health's flagship products began running three to four studies a week through Rally's MCP. “It allows more and more people to get into it with [the] rigor that we want, because you don't have to worry when someone does something that they're [going to] break a rule, because the rule is embedded," said Senior Staff Researcher Mahad Bullo.
At Calendly, first-ever Research Ops hire Ray Mason-Murillo used the MCP to make the company's existing research easily discoverable, surfacing prior interviews in seconds. "This doesn't seem like a V1 at all. This is way, way beyond an MVP," he said. Ray also noted that the MCP has quietly absorbed the operational overhead he used to carry such as naming conventions, indexing standards, and the cleanup that followed every self-serve study.
.avif)
What's generally available today
Rally Agent and MCP now cover the full arc of a study - from "who should I talk to?" through recruitment, scheduling, incentives, and post-session synthesis.
- Rally Agent: Our out-of-the-box Research Ops Agent, powering research inside the Rally app and from Slack. Tell it what you're researching and it runs the study end to end.
- Rally MCP: Rally's robust set of tools, callable from Claude or any other LLM interface your team already uses, so you can build your own agentic workflows and power ongoing research wherever you work.
Just describe what you’re researching, and Rally builds the study from your approved templates, launches recruitment from your own panel or a third-party panel, handles outreach, scheduling, and quotas, and sends incentives with automated reminders.
Here's how the Rally Agent & MCP show up across the whole study:
- Ask Rally anything. Check your schedule and who's booked, prep for an interview, or pull and analyze a transcript - all read-only, so anyone on the team can start here without changing a thing.
- Kick off research. Build interviews, surveys, and unmoderated tests from a brief or from scratch, with screeners, scheduling rules, and quotas configured for you.
- Bring in the right people, and reach out safely. Import and enrich your lists, and confirm you have enough of the participants you need prior to starting research. Send outreach with cooldown, blocklist, and do-not-contact enforced on every send.
- Rally comes back to you. The agent only pulls you in for the decisions that are actually yours - approving an outreach batch, confirming an incentive, or knowing the moment a study hits its target.
The bigger picture
Our mission is to make it easy for any product team to talk with their users. In a world where you can build almost anything, the hardest part is deciding what to build, and many of those decisions are a guess until you talk to a real person.
We've been building closely with teams at Google, Figma, Meta, Ramp, Perplexity, and more to support the future of Product Research. With general availability, it's ready for yours.
Get started
If you're a Rally customer, reach out to your CSM to discuss getting started with the Rally Agent and MCP in your workspace. If you're not yet working with Rally today, book a demo and we'll show you what agentic research looks like with your own setup.