Creating research jobs

Research jobs are the core workflow unit in Progrid AI Lead Research. Each job defines what to search for, which providers to use, and how to deliver the results. Jobs are created through a wizard that guides you through the configuration process.

Accessing the wizard

There are two ways to open the research job creation wizard:

1. Navigate to CRM ‣ AI Research ‣ Research Jobs and click the Create button.

2. From the Research Jobs list or kanban view, click the New Research Job button in the top-left corner.

The New Research Job wizard (model Progrid.research.wizard) opens as a modal dialog with all the fields needed to configure and launch a research job.

Wizard fields

Basic information

Field	Required	Description
Name	Yes	A descriptive name for the research job (e.g., “Plumbers in Austin TX” or “Competitor analysis - web design agencies”).
Target Description	Yes	A detailed description of the type of businesses to search for. This text is used to generate search queries. Be as specific as possible about the industry, location, and characteristics.
Location	No	Geographic focus for the search (e.g., “Austin, TX”, “Greater London”, “Ontario, Canada”). Leaving this blank performs a broader search without geographic constraints.
Industry	No	Industry or vertical to target (e.g., “plumbing”, “web design”, “accounting”). Helps refine search queries for more relevant results.

Provider selection

Field	Required	Description
Search Provider	Yes	The web search provider to use. Options: Brave – Uses Brave Search API Tavily – Uses Tavily Search API Mixed – Uses both providers and combines results for broader coverage
Max Results	Yes	Maximum number of search results to process per query variation. Default is typically 10. Higher values consume more API credits but may discover more businesses.

Deliverables

Configure what happens with the research results after processing:

Field	Required	Description
Create Leads	No	When enabled, results with confidence above 50% are automatically converted into CRM leads. Each lead is pre-filled with the extracted business name, email, phone, website, and a description of services offered.
Enrich Partners	No	When enabled, the system checks if a matching partner already exists (by email or website domain). If found, the existing partner record is updated with new information instead of creating a duplicate.
Export CSV	No	When enabled, a downloadable CSV file is generated containing all results with their extracted data and confidence scores.

Tip

You can enable multiple deliverables simultaneously. For example, enable both Create Leads and Export CSV to get new CRM leads and a backup spreadsheet.

Using presets

The wizard includes preset configurations for common research scenarios. Presets pre-fill the wizard fields with optimized values for specific use cases.

Contractors with outdated websites

This preset targets contractors and service providers whose websites show signs of being outdated or poorly maintained. It is useful for web design or digital marketing agencies looking for potential clients.

Pre-filled values:

• Target Description: Focuses on contractors with old or broken websites

• Search Provider: Mixed (both Brave and Tavily for maximum coverage)

• Create Leads: Enabled

Local service providers

This preset searches for local service businesses in a specific area. It is ideal for sales teams doing geographic prospecting.

Pre-filled values:

• Target Description: Focuses on service-based businesses

• Location: Must be specified by the user

• Search Provider: Brave (good local search results)

• Create Leads: Enabled

Competitor analysis

This preset is designed for market research. It searches for businesses similar to your own or your client’s business to understand the competitive landscape.

Pre-filled values:

• Target Description: Focuses on businesses in a specific industry

• Industry: Must be specified by the user

• Search Provider: Mixed

• Export CSV: Enabled (for analysis in external tools)

Note

Presets are starting points. You can modify any pre-filled field before launching the job.

Launching the job

After configuring all fields, click the Start Research button at the bottom of the wizard. This action:

1. Creates an Progrid.research.job record with status Queued.

2. Generates up to five query variations from your target description, location, and industry.

3. Enqueues the job for background processing.

4. Redirects you to the job’s form view where you can monitor progress.

Important

Jobs are processed asynchronously in the background. The page does not need to remain open for the job to complete. You will see the status update as the job moves through each pipeline phase.

Job status progression

After launching, a job moves through the following statuses:

Status	Description
Queued	Job is waiting to be picked up by the background processor.
Searching	Query variations are being sent to the configured search provider(s).
Fetching	Discovered URLs are being downloaded and content is being extracted.
Normalizing	The LLM is extracting structured business data from page content.
Scoring	The LLM is evaluating activity level and assigning confidence scores.
Storing	Results are being saved to the database.
Delivering	CRM leads, partner enrichment, and/or CSV export are being generated.
Completed	All phases finished successfully. Results are available for review.
Failed	An error occurred during processing. Check the job’s log for details.

Monitoring a running job

From the job’s form view (CRM ‣ AI Research ‣ Research Jobs and select a job), you can see:

• Current phase – Which pipeline phase is active

• Progress indicators – How many URLs have been fetched, how many results have been processed

• Log messages – Detailed processing log in the chatter, including any warnings or errors

• Result count – Number of results found so far (updates as results are stored)

Tip

Use the Refresh button or reload the page to see the latest status. The job processes in the background and the form view does not auto-refresh.