How we automated hiring triage with n8n + ChatGPT
We recently automated one of the most time-consuming non-technical processes in our company: early-stage hiring triage.
To do this, we use n8n. n8n is our platform of choice for AI automation: it connects tools and APIs, runs processes on triggers (like a new form submission), and lets us add logic, routing, and enrichment steps. We use AI assistants like ChatGPT, Claude, and Gemini daily, but n8n is the place where we orchestrate complex workflows.
Here’s what our hiring workflow looks like today.
Candidates apply through a Typeform form. The moment a submission arrives, n8n picks it up and runs an initial filter based on hard constraints. In our case, we typically hire from Spain, Portugal, and Europe, mainly for time zone alignment and cultural reasons. So applications from countries outside that scope are automatically filtered out early, before anyone spends time reviewing them.
If a candidate passes that first gate, we run an AI evaluation step. We connect n8n to the OpenAI API and use ChatGPT to review the candidate’s CV, cover letter, LinkedIn profile (if provided), and the job description. Instead of asking for generic feedback, we ask for a structured assessment: relevance of experience to the role, evidence of real production work, alignment with the required technologies, possible red flags like frequent job changes, and the overall quality and clarity of the application. The goal is not to “decide” for us, but to produce a high-signal summary that makes human review faster and more consistent.
Next, n8n creates a Linear issue via Linear’s GraphQL API. This is where it becomes genuinely useful: the candidate is automatically added to our hiring project with the right format, tagged with relevant technologies (React, Ruby on Rails, Python, etc.), and assigned a priority based on the AI assessment. The issue includes the candidate’s key details plus a qualitative AI summary so the hiring manager can sort by priority, scan the assessment, and then deep-dive only where it’s worth it.
Finally, the workflow sends an email to the candidate, so responses are timely and consistent without manual follow-ups.
The result is simple: a prioritized candidate pipeline inside the tool we already use to manage work. Hiring still needs human judgment, but the repetitive glue work around it is exactly what automation plus AI should handle.
This is just one example. We’re applying the same approach to other operational workflows: take a process that’s repetitive and context-heavy, use n8n to orchestrate the steps, and use AI where it adds leverage, then push the output into the systems your team already trusts.
If you want help designing or implementing AI-powered automations like this, GPTApps can help. We build n8n workflows, integrate AI models (ChatGPT, Claude, Gemini), and connect everything to your stack so your team spends time on decisions, not admin.
