AI Job Scraper & Resume Matcher
n8nAI-assisted workflowsStart a job search in Slack, assess openings against resume content, and prepare documents and application drafts.
- Platform
- n8n
- Focus
- AI-assisted workflows
- Evidence available
- Original screenshot
Project overview
An AI-assisted job-search workflow connects a Slack request with job retrieval, resume analysis and document preparation. The supplied updated n8n screenshot reveals the actual orchestration and a draft-only Gmail action.
Business problem
Searching for openings, checking relevance and preparing tailored application material involves repeated handoffs between chat, job listings, documents and email.
Automation solution
Validate the incoming search, retrieve jobs through the visible SerpAPI request, iterate through results, use an AI agent for resume content and match analysis, then prepare a copied resume document and route to an email draft or Slack message.
Simplified process summary. Explore the architecture below for branches and decision logic.
Workflow evidence
Original image, preserved in full. Enlarge and zoom to read individual nodes.
Workflow architecture
Trigger & input
Slack Trigger starts the visible flow. An AI step labeled “Check if the query is valid” precedes Query is Valid; the validator’s provider and exact prompt are not readable.
Processing & output
- The valid-query branch posts a “searching for jobs” Slack message, then Get All Jobs makes an HTTP GET request whose visible URL begins with https://serpapi.com. The earlier brief named JSearch; this updated screenshot supports SerpAPI instead.
- Check Results gates the response. Split Out connects to Loop Over Items, then Get Resume Content retrieves a Google document for the analysis stage. The full loop-return wiring and iteration completion behavior need the exported workflow.
- Create Resume Content and Match Analysis is an AI Agent connected to an OpenRouter Chat Model and Structured Output Parser. Its prompt, schema, scoring rules and model selection are not visible.
- Search files and folders in Google Drive is followed by a Filter, Wait, Copy Resume Template, Wait1 and Update a document in Google Docs.
- The final If node branches to Gmail Create an Email Draft or a Slack Send a message action. The capture shows draft creation, not sending a job application.
Conditional logic
Query is Valid separates accepted input from the Invalid Query Slack action. Check Results also has a fallback connection to that message. A later Filter and final If provide additional gating, but their expressions and match thresholds are hidden.
Key features
- Slack-based entry and feedback
- Query validation and result gating
- Per-item orchestration with an AI agent and structured output parser
- Resume-template copying and document updates
- Conditional Gmail draft or Slack output
Technical challenges
The visible complexity is coordinating job items, resume context and structured AI output across document and email steps. Wait nodes and conditional gates are present, but the capture does not explain their timing or establish that missing files, API limits or malformed AI output are handled correctly. No debugging history was supplied.
Business value
Intended to support a more focused job search and reduce repeated copying between tools while leaving the shown email action at draft stage. No time savings or application outcomes are verified.
Project status
Testing & limitations
Test execution shown
The original screenshot shows green execution markers and item counts on the displayed path. That is evidence of a captured test, not independent verification of output quality, all branches, continuous operation or production deployment.
Source: Kyle’s original workflow screenshot and project brief. The visible canvas has been reviewed; exported configuration, output contents and production run history have not been independently verified.
Limitations & next steps
What needs verification: resolve these open questions before treating this design as a verified implementation.
- Exact query-validation prompt, job parameters and API credentials
- AI model, output schema, matching criteria and final If expression
- Full loop-return behavior, wait durations and file-selection rules
- Draft contents, failure handling and production run history