Getting Started
Search & Prompt
Agents
Shortlist
Sequences
Contacts
Exports & Integrations
Beta Access
Organization
Support
AI Filters
Use AI filters to match all hard criteria in your search.
Getting Started
Use AI filters to effectively narrow down your search to your must-have criteria. There’s four ways to get started with AI filters:
- Prompting [recommended]
- Search by Job Description
- Boolean Filters
- Manually Select Filters
The below sections briefly cover each of the four methods, before diving deeper into the “edit filter” menu.
Prompting
80%+ of searches run on PeopleGPT use AI prompting to get started. A prompt is a description of the type of candidate you’re looking for, written out in a short paragraph or sentence.
For example, here are some well-written initial prompts:
- Marketing Manager in Europe, German-speaking, working at a large enterprise, skilled in SEO
- Senior Scientist in Australia, experience in Biotech, 10+ years of experience
- Consultant in London with 2+ years experience at top consulting firms
What makes a prompt well-written? A good prompt will explicitly describe what you’re searching for, mentioning specific requirements.
That can include fields like: location, job title, company experience, industry experience, skills, years of experience, university attended, degree held, and more.
Search by Job Description
If you prefer to search by job description, you can do so by selecting the “Search by Job Description” button. To enter your job description, you can choose from two methods:
- Upload: upload a PDF, plain text, Word document, Google Doc, or similar.
- Text: copy-paste from your source. No need to adjust formatting.
PeopleGPT will then configure a search query for you using your job description. It will analyze suggested filters to match by, which can still be adjusted afterwards, too.
Boolean Filters
Do you have pre-configured search strings? Use the Boolean Filters option to kick-off a PeopleGPT search with your search string.
Select Manually
Selecting filters manually advances you to the select filters page directly, allowing you to configure your search from scratch. In the majority of cases, using the prompting method will help you set up your filters more quickly.
Editing Filters
Regardless of which method you use to get started, you can use the Edit Filters page to dive deeper.
Overview
Select the “Edit Filters” button to open the menu. You can review and adjust the search filters as needed.
Smart Search Fields
Many of the edit query search bars are smart AI search fields. That means you can enter either a specific entity, or describe what you’re searching for. For example, in the companies search bar, you could enter:
- A specific company: “Google” → Select Google from the dropdown menu
- A description of companies: “energy companies in New York” → Select the “Ask AI” field to activate the smart AI search
Smart search fields aren’t just limited to companies! Try them in other fields, too, like Company Industries or Universities.
Pinned Skills
By default, PeopleGPT treats skills requirements differently depending on the number of skills entered:
- When you enter a single skill, the system will treat it as a must-have. Only candidates with that skill will appear.
- When you enter multiple skills, PeopleGPT will show results that meet at least one of the required skills.
In cases where you have multiple must-have skills, you can use the pinned-skill icon to make the specific skill mandatory. In the example below, the search will run as follows: show candidates that are skilled in Python (mandatory) and skilled in Leadership or Communication (one of both).
Excluded Fields
PeopleGPT has a multitude of smart search fields — including expanding job titles to help find best fit candidates. For example, entering “Software Engineer” will also help you find Principal Software Engineers. However, in some cases, you may want to restrict how broad the search is.
You can solve this by using our “Exclude” field, which is a stop icon that appears when hovering over a job title, skill, industry, or location. In the search below, we are searching for:
- Location: Someone in California, who is not in Los Angeles
- Job Title: Software Engineer, who is not a Software Engineering Manager
Additional Filters
Advanced company filters allow you to construct complex searches based on data such as company funding stage, funding amount raised, investors, and revenue. They work the same as other filters, but contain some additional configuration options.
Investors, Funding, Revenue, and Funding Stage
You can combine a variety of advanced fields to power even more complex searches, such as profiles at a company with $25-$100m in revenue, that has raised $10-$50m in funding, and has worked at a company backed by Sequoia Capital.
If you included a specific field in your initial prompt to trigger advanced filters—like 'Software Engineers at Series A startups'—those filters will already be applied.
Company + Funding Stage
Want to find profiles who were early at certain companies? Enter your preferred companies, select the “Funding Stage” drop-down, and configure your desired funding stage. This will show you profiles that were at your selected companies at a specific stage.
Power Filters
PeopleGPT also provides access to Power Filters, which can be used in combination with other search fields. The drop-down at the top allows you to select between “Match Any” and “Match All”, which will treat your power filter selections as “OR” or “AND” searches, respectively.
Power filters are split into three primary categories: Diversity, Career Growth, and Achievements. They are continuously expanded. If you have additional requests for power filters, please submit them here.
Likely to Switch
The 'Likely to Switch' feature offers six different indicators to help assess how open someone may be to new opportunities. You can toggle these signals individually or enable them all at once— perfect for zeroing in on candidates who are most likely to respond with interest. If you hover over any of these with your cursor, you'll see a quick explanation of what AI is doing behind the scenes to generate each signal.
← Previous
Next →