AI vs. AI: Snuky’s New Look

Snuky in a Nutshell

I started working on Snuky a week or two ago. I mentioned it in my last article (How AI Is Driving Your Job Search Without You Knowing) where I discussed using AI tools to "beat the system" (the system being any AI-driven ATS).

Snuky is my AI-driven platform for automating job searches and applications. Tracking them in Google Sheets was burning me out, so I decided to code something to solve that pain point in my job search. As of today, Snuky is tracking my applications, monitoring job postings, and notifying me of changes to job postings. So far it's made job searching easier for me—especially when it comes to notifying me about job postings being removed.

Looking Forward

Job Description Matching

By reverse-engineering AI models and ATS keyword matching (NLP-driven), I think I've figured out how they score resumes against job descriptions—they (or recruiters themselves) reject candidates that don't meet a specific threshold match (e.g., 50%). Snuky does well with extracting job description keywords, but calculating my resume's exact match score is iffy since AI models define "matches" differently. Using this knowledge, I'll combine the results of different AI models and use the average match score.

More Automation

This is just the beginning of my automated job search process. I'm planning to automate it to the point I don't have to search for jobs. My plan is to have Snuky search for jobs (specifically job descriptions on job boards) that match my resume or could match my resume (based on keyword matching). From there it will take my resume and tailor it to the job description. Currently, that's a manual process for me and it's tedious AF.