Rose AI Search
Brief Rose the way you'd brief a recruiter. She asks what she needs, searches, and brings back real candidates for you to judge.
Rose AI search is the fastest way to find candidates. Rather than setting filters by hand, you describe who you're looking for in plain language — the same way you'd brief a recruiter — and Rose does the research.
The Rose home screen
Click Rose in the sidebar (or New chat). You'll see “Hi, how can I help?” and a message box. A few things live around it:
- Working with — your company context from the Knowledge Base, with an Edit link.
- Project — choose Start a new project or pick one of Your projects, so the candidates you find land in the right place from the first message.
- Mode — Planning or Normal. See Planning Mode.
- Paperclip — Upload a job description (PDF, DOCX or TXT) and Rose reads it for you.
- Exclude — keep people you've already sourced out of the results.
- Browse skills — see what Rose can do, with example prompts.
Briefing Rose
Type a description of the role and the person. Be as specific or as broad as you like:
“I'm looking for an Account Executive with 3–5 years of experience at a New York tech startup”
“Find me a senior DevOps engineer in London or Manchester with AWS and Kubernetes experience”
You can name several cities in one brief. Rose treats Remote as a way of working, not a place, so she won't search for people who happen to live “in Remote”.
Rose asks what she needs
Before she searches, Rose makes sure she has the essentials. When something is missing she asks — often several questions at once on a single card, with suggestions you can tap and room to type your own answer. Answer in any order; Rose keeps track of what she still needs.
Before her first search, Rose asks for the job description. Paste it, upload it — or, if you haven't written one yet, choose Draft with me and she'll build it with you from a few answers.
How Rose searches
Rose picks the right approach for the brief:
For most roles, Rose shows you three profiles at a time and learns from your verdicts until the search is right. This is the most precise way to search. See Calibrating a Search.
When you already know exactly who you want — named companies, a specific title — Rose searches live and groups the results into named clusters you can review and add from.
Describe the kind of company you want to hire from and Rose finds them first — competitors, adjacent markets — then sources from the ones you pick.
Rose can also search the web while she works — to look up a company, a product or a market — and shows you what she found.
The experience number is smarter than it looks. Rose recomputes each candidate's relevant years of experience — counting only post-graduation, in-function work and excluding internships, student jobs and unrelated roles.
Adding candidates to a project
Rose never adds anyone to a project without your say-so. Candidates join a project when you mark them a good match during calibration (with Add liked candidates to project switched on), or when you ask Rose to add them. Once they're in, you'll find them on the project's pipeline — see Managing a Project.
Refining in conversation
Because the search lives in a chat, you can steer Rose like a colleague — narrow by location, seniority or companies; cap how many candidates come from one employer; set a minimum experience bar; or tell her what you liked and didn't. Every chat is saved under Recent in the sidebar, so you can come back to it later.
Rose is AI and can make mistakes. Always review profiles before you launch outreach.