The User Research Ally is 12 specialist AI agents covering every stage of your workflow — trained by a researcher with 12+ years in the field, not an AI with no idea what a double-barrelled question is.
Most researchers aren't struggling because they lack skill. They're struggling because they're under-resourced, under-time, and carrying the entire methodology in their head with no one to gut-check them.
No senior researcher to sanity-check your screener. No one to catch the leading question in your guide before the session. You're the first and last line of quality — and that's exhausting.
Research plans from scratch. Guides re-written every time. Insights you meant to write up "later" sitting in raw notes from three sprints ago. The work piles up faster than you can clear it.
Fourteen stakeholders with fourteen opinions on what the research should find. A roadmap that doesn't wait for your insights. And a blank doc every time you open a new project.
The User Research Ally lives in your browser. Every agent is waiting for you the moment you log in — no setup, no re-explaining your methodology, no starting from scratch.
Choose monthly or annual. You'll get a link to access your research portal — a private, password-protected space with all 12 agents ready to go.
Takes 2 minutes to set upEach agent asks you questions first — your project context, your constraints, your research question. It gathers what it needs, then hands you something finished.
No prompt engineering neededA complete research plan. A guide ready to facilitate. Insights that have already been checked against your business objectives. The kind of output that used to take hours.
250 uses every monthThe User Research Ally runs in your browser via a secure portal. You'll create a free Pickaxe account to access it — takes 60 seconds. No app to download, no installation, works on any device including mobile.
Each time you send a message to an agent, that's one use. A typical full agent session (back-and-forth to a finished output) uses 10–15. In practice, 250 uses covers roughly 15–20 complete agent sessions per month — more than enough for active research work. Resets every month.
Think of each agent as a knowledgeable colleague sitting next to you. They ask questions, catch mistakes before they happen, and hand you something finished at the end.
Walk into every study knowing exactly what you're trying to find out, how you're going to find it, and what success looks like. No more "wait, what's the actual research question here?" at 11pm before your first session.
Sit down to facilitate any usability session fully prepared: tasks written, prompts ready, facilitation notes in hand. The kind of prep that used to take half a day, done in minutes.
Run discovery interviews that actually surface what's true, not just what people think you want to hear. (BTW "what features would you like?" is not a research question. This agent knows that.)
Get brutally honest feedback on your ideas before your team spends three sprints building the wrong thing. Real reactions. Real insights.
Stop talking to the wrong people. A screener that filters for your exact user, so your research reflects reality — not just whoever was willing to hop on a call.
Build surveys that produce data worth showing to a stakeholder. No leading questions. No double-barrelled items. No accidentally "proving" whatever you already believed going in.
Turn a mountain of notes into clear, evidence-backed insights your team will actually act on. This is the part everyone skips or does badly. Not anymore.
Plan your research programme across a quarter or a product cycle. Right methods, right timing, right questions. Stops research from being reactive and one-off.
Build a research roadmap that actually makes sense alongside your product roadmap. Know what you need to learn, when you need to learn it. No more research that shows up too late to matter.
Maps your research questions directly to business objectives, team goals, or OKRs. Makes it impossible for stakeholders to ask "why are we doing this study?" And makes prioritisation a lot easier.
Paste in your survey results and get back something actually useful. Patterns, distributions, what the numbers mean in plain English. Because collecting data is the easy part.
Right after a session, while it's still fresh: capture key observations, notable quotes, and emerging themes in minutes. Stops "I'll write it up later" from becoming "I can't remember what they said."
This is the User Research Plan agent — one of 12 in the full subscription. Describe your project and watch it ask the right questions. No account needed, no credit card, no strings.
Free trial · Usage limited · No account required
Walk into every study knowing exactly what you're trying to find out, how you're going to find it, and what success looks like. This agent asks questions about your project, your team, your constraints — then hands you a complete research plan.
First 50 spots only · Price locks in forever
Every subscription comes with two additional products at no extra cost. Not stripped-down versions. The real things.
My complete AI prompt library — hand-crafted for researchers. Synthesis, analysis, writing, facilitation, stakeholder communication. Every prompt I actually use. Not generic AI suggestions padded to look impressive.
52 research frameworks installed directly into Claude.ai — TEDW, Pyramid Principle, insight formula, ROI metrics, and more. Install once, and Claude applies them automatically every session. Works alongside the Ally, not instead of it.
This isn't a prompt wrapper. It's a thinking partner — trained on 12 years of real UXR work — that walks you through every problem from first brief to final readout.
Every agent is trained on real source material — not generic AI knowledge. Real case studies from real companies. Real screeners that actually recruited the right people. Real insights that actually changed roadmaps. Real frameworks refined over hundreds of research sessions.
Every agent starts by asking you questions — about your project, your constraints, your stakeholders, your timeline. It gathers context before it builds anything. The output reflects your situation, not a generic template. This is what makes it an ally rather than a search engine.
The 11pm blank doc. The stakeholder who wants a survey for everything. The insight that doesn't quite land. The research plan for a study you've never run before. The Ally is there for exactly those moments — not to replace your judgment, but to make sure you never have to figure it out alone.
Fair question. Here's the honest answer — and why most researchers who try generic AI tools end up frustrated.
| What you get | ChatGPT / Claude | Dovetail / Maze | User Research Ally |
|---|---|---|---|
| Knows what TEDW questions are | ✗ Needs explaining every time | ✗ Not AI-assisted planning | ✓ Built in from the start |
| Covers the full research lifecycle | ✗ Generic, no UXR structure | ✗ One or two phases only | ✓ Planning → Synthesis → Roadmap |
| Understands research quality standards | ✗ Will happily write leading questions | ✗ No methodology enforcement | ✓ Catches bad questions automatically |
| Connects research to business goals | ✗ Not without a long prompt | ✗ Not a feature | ✓ Research to Goals Mapper |
| No re-explaining your methodology | ✗ Every session, from scratch | ✗ N/A | ✓ Already knows how you work |
| Monthly cost | $20/month | $35–$199/month | $39/month (founding) |
"I ran a full generative study last sprint and didn't start a single document from scratch. The Research Plan agent asked exactly the right questions. The Interview Guide caught three leading questions I would have missed. This is what having a senior researcher next to you feels like."
"The Insight Builder is the one I use every single week. Key learning + why + consequence, enforced every time. My last three readouts had action items assigned in the room. That never happened before I started using this."
"Solo UXR with no one to gut-check me. The Research to Goals Mapper made my last research proposal impossible to reject — I walked in with a document that connected every study question to a team OKR. It went through in one meeting."
Founding member pricing is locked in for as long as you stay subscribed. After the first 50 members, it goes to $49/month.
Yes — you'll need a free Pickaxe account to access your research portal. It takes about 60 seconds to create and just needs an email address. No credit card required for the account itself — your subscription payment goes through Stripe and then unlocks access.
Once you subscribe, you'll get a link to your private research portal. Log in with your Pickaxe account and all 12 agents are right there — each one opens a chat interface in your browser. No app to download. Works on desktop, tablet, and mobile.
Each time you send a message to an agent, that's one use. A typical full agent session — where the agent asks questions and builds your output — uses around 10–15 messages. So 250 uses covers roughly 15–20 complete agent sessions per month. For most researchers doing active project work, that's generous. Resets every 30 days.
You won't be able to use the agents until your usage resets on your monthly billing date. You can track how many uses you have left inside your portal at any time. In practice, most researchers use 50–120 per month — 250 is a genuinely generous cap for real research work.
This is the right question to ask and you deserve a detailed answer. Here's exactly how data flows through the User Research Ally:
The platform (Pickaxe): The User Research Ally runs on Pickaxe, which is SOC2 Type II and GDPR compliant. Pickaxe does not use your conversation data to train AI models. Your conversations are not shared with other users or third parties for commercial purposes. Pickaxe's full privacy policy is available at pickaxe.co/privacy.
The AI model: Agent responses are processed via Claude (Anthropic). Anthropic's API does not use API-processed conversations to train models by default — this is distinct from the Claude.ai consumer product. Your inputs are processed to generate a response and are not retained for training purposes.
What this means practically: Your research questions, your study designs, your methodology notes — none of this is used to train AI or shared outside the platform. Conversations are stored to display your history within the portal, but are not mined for commercial purposes.
What to avoid: As with any cloud-based AI tool, avoid sharing raw participant data (recordings, transcripts with real names), personally identifiable information about research participants, or anything your organisation classifies as confidential under its own data policies. Use the Ally for methodology, structure, and synthesis — not as a data storage system.
For enterprise or regulated environments: If you work in healthcare, finance, or other regulated industries with strict data handling requirements, check your organisation's policies before using any AI tool with work-related content. When in doubt, anonymise before you share.
Generic AI tools don't know what TEDW questions are. They'll happily write a leading screener question and never tell you. The User Research Ally is trained on 12 years of real UXR methodology — it knows the frameworks, catches the mistakes, and structures outputs the way research actually works. You don't have to explain anything.
Good news: you get both. The UXR Claude Skills Bundle (normally $49) is included free with every User Research Ally subscription. They're designed to work together — the Skills Bundle installs research frameworks into Claude.ai so Claude responds the right way when you prompt it directly. The Ally is a fully built agent system where each agent asks you questions and produces finished outputs without any prompting needed. Skills make your Claude smarter. The Ally does the work with you. Together, they cover everything.
No. The User Research Ally runs entirely through its own portal — you don't need any other AI subscription to use it. It works independently of Claude.ai or ChatGPT. The only account you need is the free Pickaxe account to log in.
Yes, anytime, no questions asked. Your access continues until the end of your current billing period. If you're a founding member and cancel, your founding member price is gone — re-subscribing later would be at the standard price.
Yes — and this is one of the strongest use cases. If you're a researcher responsible for enabling PMs, designers, or CSMs to run lightweight research themselves, the Ally gives them a structured process to follow rather than a blank prompt box. The methodology guardrails are built into every agent — so a non-researcher using the Screener agent can't accidentally build a demographics-first screener, and a designer using the Survey Builder will get flagged on leading questions automatically. You get scale without losing quality control. Your role becomes oversight and strategy rather than fixing everyone's methodology at midnight.
This is for practising UX researchers — solo researchers, small teams, and research-adjacent practitioners who already understand research methodology and want AI that matches their level of rigour. It is not a course or introduction to research. If you're already doing research and want to do it faster and better, this is for you.
12 specialist agents. 250 uses a month. 90 expert prompts + the full Claude Skills Bundle — $198 of extras, included free. The colleague you never had, available 24/7.