Show HN: Good old emails and LLMs for automating job tracking

by adelowo on 6/21/2025, 9:37:48 PM with 5 comments
So I spent the last few days building Jobstack. The logic is quite simple. You apply to jobs and you get emails, you trade emails back and forth from interviews, questions and others until the role is either accepted or you are rejected.

Also easy to apply to hundreds of roles and not being to know where you stand easily. With Josbtack, you sign up, get a unique email and forward emails to the url. And it uses LLMs to extract company details , tries to find information online about them and presents that to you.

Every email you forward becomes part of your timeline with the company.

It also tracks rejection, offers from the emails too and gives you a nice stats dashboard amongst others.

Using Gemini 2.5 pro right now. No data stored not in any way. After extraction, it’s discarded. Even “AI chats with the company” aren’t stored

https://jobstack.me

  • by Yuki_Nishikawa on 6/21/2025, 11:20:59 PM

    This scratches a very real itch. I've hit the "can't remember which recruiter said what" wall multiple times, especially when applying in bulk.

    A couple of questions: - How reliable is the rejection/offer detection logic? Are you using simple keyword matching or something more contextual with Gemini? - How do you handle company name normalization across emails? (e.g., "Google LLC" vs "Google Careers")

    Also curious: any plans to expose the timeline as a downloadable JSON or API? Would be neat for power users to integrate into personal dashboards.

    Great job building this — surprisingly underserved use-case despite how common it is.

  • by remram on 6/21/2025, 9:40:00 PM

    So you are collecting sensitive, valuable data just to save me from checking boxes in my list myself?