Job Application Tracker

Job hunting is admin-heavy: 50+ applications, status changes weekly, follow-ups due at specific times, and you need to remember the company basics before each interview. Most people start with a Notion template, give up after 2 weeks, and live in browser tabs. The agent absorbs the admin so you focus on the actual conversations.

What it does

Quick capture: "applied to <company> for <role>" → logged
Tracks status: applied → screen → interview → offer / declined / ghosted
Follow-up reminders: 7 days after applying without response, prompts a follow-up
Interview prep: pulls company recent news, your contacts there (from Gmail), prior application notes
Weekly summary: where things stand
Post-interview note capture: voice or text → saved tagged

What you'll need

Memory for application records
Gmail for recruiter email integration (optional but useful)
Cron for follow-up reminders + weekly summary

Setup

1. Define the record shape

Send to Flowly

Remember the job application schema: Each application: company, role, status, applied_date, source (LinkedIn / referral / direct / etc), recruiter (name + email if known), salary (if known or asked), location/remote, notes (free text), updated_at Stages: - applied: just submitted - screen: phone/email screen scheduled or done - interview: technical/onsite scheduled or done - offer: received - declined: by them OR by me (annotate which) - ghosted: 30+ days no response - closed: I withdrew Tag every record "job:<company-slug>".

2. Quick capture conventions

Send to your bot:

"applied to OpenAI for Founding Engineer, Remote — via LinkedIn, referred by Erkan"

Agent parses and creates the record. Future updates by reference:

"OpenAI screen scheduled for Wed 2pm with Sarah" → status=screen, adds note "Sarah, Wed 2pm"

"OpenAI moved to interview, technical Thursday" → status=interview

3. Follow-up cron

Send to Flowly

Cron "job-followups" daily at 10 AM: For each application with status="applied" and applied_date > 7 days ago and last_update <= applied_date: Prompt me in Telegram: 📨 Follow-up time: <Company> (applied 8 days ago, no response yet) Want me to draft a follow-up email? Reply yes / skip / mark ghosted. For each with status="interview" or "screen" and the scheduled date is tomorrow: 📅 Tomorrow: <Company> <Stage> at <time> Want a prep brief? Reply yes / no.

4. Pre-interview brief

Send to Flowly

When I say "prep <company>": 1. Pull the application record 2. Search my Gmail for any thread with @<company> domain in last 60 days (recruiter conversations have useful context — did they mention anything about team or interviewer?) 3. Web search for company recent news (funding, releases, hires) 4. memory_search for "company:<slug>" in case I have prior notes (e.g. friends who worked there) 5. Output: - Role & responsibilities (from my notes) - Recent company news (3 bullets) - Interviewer context (if any) - Likely questions for this stage (e.g. for a screen: motivations, basic background) - Open questions I should ask them (3 generated based on stage)

5. Weekly summary

Send to Flowly

Cron "job-week" every Monday at 9 AM: Summarise: - Applications by status (counts) - New movement this week (advances, rejections, offers) - Pipeline velocity (avg days from applied → interview) - Top 3 most-promising (latest activity, status >= screen) Send to Telegram.

Tips

Don't try to capture in real-time. Apply on LinkedIn, copy the URL, paste to bot ("applied to <X> via <URL>"). Two seconds.
Tag every email. When a recruiter emails, tell the bot "applied at OpenAI — recruiter just emailed Mira about screen next week." The Gmail link picks up; the bot updates status.
Prep is where this earns its keep. Walking into an interview having read the latest news and your contact's notes is the highest-ROI 5 minutes you can spend.
Status decay is a feature. Anything stuck at "applied" for 30 days = ghosted. Mark and move on. Don't romanticise.
Salary info matters. Even if asked late, log it. Patterns across companies inform negotiation later.
After landing, archive. Don't keep your job search history forever. Memory tag with archived after the new job starts; the agent stops surfacing them.