Run real research from any tool you already work in.
Connect Alchemic's MCP server to your agent, app, or assistant and trigger a full study end-to-end — guide, fielding, AI-moderated interviews, cited findings. Run it on your own customer list, or recruit fresh respondents from Alchemic's panel. No dashboard to open.
The question shows up in your workflow. The research happens somewhere else.
Your team lives in Slack, Microsoft Teams, a notebook, an internal agent. But the moment you need to hear from real people — your own customers or the wider market — you leave. You brief an agency or a panel partner, wait weeks, then stitch the output back in by hand.
An AI agent can summarise what you already know. It can't go ask two hundred real people a question and come back with cited answers. That gap is where decisions stall — and it's exactly the gap the Alchemic MCP closes.
A full study, callable from wherever you work.
Trigger a full study with one call.
Point your MCP client at Alchemic and describe the objective in plain language. Alchemic writes the discussion guide, fields the study, moderates every interview, and analyses the results — the same engine behind our managed studies, now callable from the tool you're already in.
Your own customers, or the open market.
Field to a list you bring — CRM exports, churned users, NPS detractors, a support cohort — or recruit fresh respondents from Alchemic's own panel when you need non-customers. Same workflow, same call. You just choose the sample.
- CRM & CX exports
- Churned / lapsed users
- NPS detractors
- Your respondents, your consent
- Recruit non-customers
- Category & competitor users
- Screened to your spec
- Incentives handled
sample: "my_list" | "panel"Cited findings, back in your tool.
Themes, verbatim quotes, segment cuts, and respondent counts come back into the same conversation — every claim drillable to the people behind it. Pipe them into a deck, a Slack or Teams thread, or the next step of your agent.
Every line traces back to a real interview — drill from theme to quote to the respondent who said it.
From objective to cited findings, hands-off.
Connect the MCP server.
Add Alchemic to your MCP client — an AI assistant, a coding tool, or your in-house agent. One config, authenticated to your workspace. Every Alchemic capability becomes a tool your agent can call.
Describe the study, choose your sample.
Say what you want to learn and who from. Bring your own respondent list or let Alchemic recruit from its panel. Alchemic handles the guide, screener, quotas, and incentives.
Field, moderate, analyse — automatically.
Alchemic runs the AI-moderated interviews over WhatsApp, web, or phone, flags low-quality responses, and returns cited findings as the study lands. No analyst on QC, no ops team chasing payouts.
Built different, by design.
Real interviews, not synthetic answers.
Every study fields actual AI-moderated conversations with real people — your customers or panel respondents. The moderator probes hesitation, follows up, handles voice notes, and works across languages. Not a model guessing what someone would say.
Bring your data, or use ours.
Run on a list you own, or recruit fresh from Alchemic's panel — from the same call. Most tools make you pick one world. This one works in both, so a customer deep-dive and a market read share a single workflow.
The whole operation, absorbed.
Screening logic, quota management, incentive payouts, dropout detection, response-quality flagging — handled by the platform and exposed as a handful of simple tools. No panel ops, no spreadsheets, no analyst spending half the week on quality control.
Your insights compound.
Every study runs against your persistent research history — category vocabulary, brand context, prior waves, segment history. Each round comes back sharper and more on-brief than the last. Legacy stacks start every project cold; this one builds on itself.
Real studies your team is already asking for.
Churn & retention deep-dives
Field to your own lapsed and churned users to find the real reason they left — not the box they ticked on exit.
Concept & product reads
Recruit fresh panel respondents to react to packaging, pricing, or a new idea against the category leader.
Continuous CX pulse
Point a recurring study at your support or CRM cohort and let findings stream into Slack, Teams, or WhatsApp.
Category & competitive intelligence
Use Alchemic's panel to hear from non-customers and switchers across the category, on demand.
Frequently Asked Questions
What is the Alchemic MCP?
It's an MCP server that exposes Alchemic's research platform as tools any compatible client can call — so an agent or app can launch and analyse a full study without anyone opening a dashboard. Define the objective, choose the sample, and the study designs, fields, moderates, and returns cited findings on its own.
Which tools can connect to it?
Any MCP-compatible client — AI assistants, coding tools, and in-house agents among them. If your tool speaks the Model Context Protocol, it can call Alchemic. Connection is a one-time config authenticated to your workspace.
Can I run a study on my own customers?
Yes. Field to any list you bring — customers, churned users, a CX cohort — or recruit non-customers from Alchemic's panel. You choose the sample source per study, from the same workflow.
How fast do studies come back?
Days, not weeks. Fielding and analysis run continuously, and findings stream back as responses land — so you see the picture forming rather than waiting for a final report.
Is my customer data safe?
Your respondent lists stay inside your workspace and are used only for the studies you run. Full security and data-handling details live on our trust page, linked in the footer.
How is it priced?
Per study, by sample size and source — running on your own list costs less than recruiting from the panel. Book a demo for a scoped quote against your use case.
How AI-driven research actually works.
What is an MCP?
The Model Context Protocol (MCP) is an open standard that lets AI applications connect to external tools and data through a common interface. An MCP server exposes capabilities — like running a research study — that any compatible client can call directly, so the AI can act in the real world rather than only answer from what it already knows.
What is a market research MCP?
A market research MCP is a server that lets an AI agent or app run real research through standard tool calls — designing the study, interviewing real people, and returning analysed findings — instead of a person operating a research platform by hand. Alchemic's MCP does this end-to-end, on your own customers or a recruited panel.
Can I run market research from an AI agent?
Yes. With Alchemic's MCP server connected, an agent can launch a full study, field it to your customers or a panel, and get cited findings back — all from inside the tool you're already working in. The agent issues the calls; Alchemic does the interviewing, quality control, and analysis.
How is this different from just asking an AI agent?
Asking an agent gives you a summary of what's already known. Running a study through Alchemic's MCP collects new primary data from real people, then returns findings with quotes and respondent counts behind every claim. One reorganises existing knowledge; the other goes and creates it.
Can I research my own customer list instead of a panel?
Yes. Point the study at a list you own and Alchemic moderates interviews with those people directly. When you need non-customers — switchers, category users, a fresh read on the market — the same workflow recruits from Alchemic's panel instead.