Run LinkedIn or Indeed searches and let GetMe Hired collect the roles, descriptions and key details for you.
No jobs waiting.
2
Review
Yes
0No
0
Stack
Tech stack
Main app components and responsibilities.
Backend
- Python 3: app logic, job collection, parsing and local APIs.
- Flask: web server, page routing, JSON endpoints and settings management.
- SQLite: local/dev database for jobs, searches, chats, settings and saved keys.
- Authlib OAuth: Google login with Flask sessions and temporary development access.
- Requests: LinkedIn HTML retrieval through accessible public endpoints.
- BeautifulSoup: HTML parsing, description cleanup and requirements extraction.
- Thread worker: runs searches in the background without blocking the interface.
Browser automation
- Chrome extension: bridge between the web app and Indeed in the user’s real Chrome session.
- User browser session: Indeed is read without managing passwords or centralized logins.
- Cooperative stop: the web app can ask the extension to stop between jobs and pages.
- Playwright: not used in the current Indeed flow; kept only as a legacy/dev dependency.
Frontend
- Server-rendered HTML: main Search, Jobs, Cover Letter and API layout.
- Custom CSS: responsive interface, side panels, job cards and Yes/No lists.
- Vanilla JavaScript: UI state, API calls, autosave, previews and job review.
AI and user data
- OpenAI, Gemini Flash or Gemma: cover letter generation from the full job description.
- Gemma: planned option for the final app through a compatible API service or local/server runtime.
- About You and Cover Style: CV, tone, instructions and preferences saved per user.
- Requirements/Filters keywords: editable lists for requirement extraction and exclusion signals.
Future deployment
- GitHub: private repository, branches, commits and controlled release flow.
- DigitalOcean: web app hosting, environment variables, HTTPS and connected domain.
- PostgreSQL: next step for robust multi-user production data separation.
- Stripe: subscriptions and payments for the commercial version.
A
Admin Dashboard
Accounts, usage and early access monitoring.
Accessi
Accessi
Only approved Google emails can sign in. Paolo remains super admin.
Users
Accounts
People who signed in with Google.
Runs
Recent Searches
Latest collection sessions and imported job counts.
API
API Calls
AI calls from cover letters, chat and internal requirements guessing.
Auto
AUTOGUESS
Contratto operativo per mostrare solo requirements verificabili nella Jobs card.
Garanzia reale
- La Jobs card non deve fidarsi direttamente del campo requirements salvato.
- Ogni card passa dalla route reale /api/jobs/review, dentro job_payload.
- Il testo viene normalizzato, deduplicato, pulito e poi validato da un quality gate finale.
- Se il gate non trova requirements affidabili, la card deve mostrare No clear invece di inventare contenuto.
Ordine di scelta
- Full description: la preview/description completa è la fonte principale, non uno snippet corto.
- Lista heading: cerca sezioni ricorrenti come Requirements, Qualifications, Skills, What you bring.
- Parser locale: divide la description in sezioni e seleziona solo blocchi rivolti al candidato.
- Autoguess/GPT: può aiutare quando il parser non basta, ma il risultato torna sempre al gate finale.
- No clear: è corretto quando non esiste una sezione chiara o la description è incompleta.
Passa se contiene
- Esperienza richiesta, anni, seniority, degree, qualifiche o certificazioni.
- Skill tecniche, tool, lingue, portfolio, right to work o disponibilità richiesta.
- Frasi rivolte al candidato: hai, sei, possiedi, conosci, must have, you have, you bring.
- Liste concrete di criteri di selezione, anche se il titolo della sezione è generico.
Non passa se contiene
- Benefits, salary, workplace, job type, contract details, hiring process o application questions.
- About us, company intro, values, culture, why join us, cosa offriamo o perché unirti.
- Responsabilità generiche, missione del ruolo, job overview o cosa farai senza criteri candidato.
- JSON, errori API, testo tecnico, duplicati, footer, privacy, equal opportunity o testo legale.
Quality gate finale
- Il gate può bocciare anche requirements già salvati se sembrano benefit, intro o mansioni.
- Prima prova a recuperare una sezione migliore dalla description completa.
- Se il recupero non passa lo stesso gate, la card diventa No clear.
- Meglio una card vuota e controllabile che una card sbagliata.
Fallback visibile
Requirements not found in a clear section.
- Non mostra JSON, tracebacks o errori tecnici al cliente.
- Gli errori API/rete diventano messaggi umani, non contenuto della card.
- Se la description del portale manca o è troppo corta, non si inventano requisiti.
- La preview resta il riferimento per decidere manualmente.
Review manuale obbligatoria
- Apri Preview: confronta la card con la full description dello stesso job.
- Trova l'heading reale: requirements, qualifications, skills, what you bring o equivalente.
- Classifica l'errore: corretta, troppo larga, benefit, mansioni, intro azienda, duplicata, low quality o No clear mancante.
- Promuovi heading: aggiungi alla lista centrale solo pattern ricorrenti e verificati su più portali.
- Non migrare dati: le card vecchie si rivalutano a render-time; riscrivere il database richiede un'azione esplicita.
No-clear
Requirements No-clear
Paid requirements calls that did not produce a useful card section.
Guess
Requirements Guesses
Suggested headings from users, ready to copy into the central requirements engine.
Inbox
Contact Messages
Access requests, feedback and early beta messages.
Instagram Feed
Upload, reorder and publish the posts shown on the home page.
Upload
Posts
Bulk upload images, then move them into the exact order you want.
22:27
5G
?
Tutorial
Set up keywords, About You, Cover Style and review flow so the app shows the right information fast.
Filters
Exclude faster
These are the keywords that may exclude you from an application and may appear in your niche. Go to Filters on the home page to edit them.
Flag highlights jobs that contain those exact words anywhere in the job text, so you can review them faster. Exclude moves matching jobs directly to No.
When in doubt, start with Flag, test the results, then switch to Exclude only after you know the keyword is reliably not relevant.
Filters
x
Flag
Exclude
Requirements
Show what matters
The AI scans each job description, matches the right requirements section, and shows it as the first thing in the Jobs card so you can decide fast.
Digital Marketing Strategist
Milano, Italy
intern
✓
Requirements found
Shown first in the review card.
Requirements
- 3+ years of experience managing paid and organic campaigns
- Strong understanding of SEO, analytics and social media channels
- Ability to turn performance data into clear recommendations
LinkedIn
Preview
System Prompt
Candidate profile
This first section should contain all candidate information: CV, work experience, education, skills, goals, tone and personal positioning.
CANDIDATE INFORMATION Name: Current role / profile: Professional summary: Education: - Work experience: - Core skills: - Projects and tools: - Languages: - Career goal: -
System Prompt
AI writing instructions
The second section should instruct the AI as a writer: tone, structure, phrases to avoid, length and final output rules.
AI WRITING INSTRUCTIONS Your task is to write a cover letter for a job application using the information provided. What you MUST do: - Write in a natural, human tone. - Keep it simple, direct and clean. - Focus on how I create value for the company. - Use real experience, not generic claims. - Show thinking, not just a list of skills. - Use a slightly confident tone, not arrogant and not weak. - Make it sound like a builder, not a talker. What you MUST avoid: - Do not use generic corporate cliches. - Do not use long sentences. - Do not repeat the same concept. - Do not include address, email, phone number or date. - Do not add notes, analysis, markdown or bullet points in the final cover letter. Structure: - Start with value. - Mention relevant experience shortly. - Explain how I think and work. - Add an AI or automation angle only if relevant. - Close cleanly. Final rule: If it sounds like AI wrote it, rewrite it. Start with Dear Hiring Manager or Dear [Company] team.
Workflow
How to use the app
1
Search
Choose the source, add the role and location, then start.
New search
in LinkedIn
i Indeed
software engineer
london
2
Jobs
Review one job at a time. Requirements and filters guide the choice.
Acme Inc.
Product Manager
Remote
Key requirements
- Product roadmap ownership
- Analytics and stakeholder work
3
Cover Letter
Open Cover Letter from a saved job. The chat starts with context ready.
Saved
Cover Letter
Write or generate a cover letter...
↑
4
Preview and Applied
Preview opens the saved job description in-app. Applied is a manual grey/green toggle.
Tesla
Reinforcement Learning Engineer
Palo Alto, CA
Indeed
Tesla
Reinforcement Learning Engineer
Palo Alto, CA
x
Job description
Build and improve learning systems for robotics teams. Work with data, simulation and model evaluation to improve real-world performance.
Requirements
Experience with machine learning workflows, Python and cross-functional product teams. The full saved job description appears here without leaving the app.
5
Stats
Stats shows found, discarded, relevant and applied jobs. Reset Numbers clears stats only. Reset Memory removes saved jobs and seen memory.
Stats
Stats
x
Jobs Found428
Jobs Discarded386 (90%)
Relevant Jobs42 (10%)
Review Time21:34
Applied14 (3%)
Total Time30:20
Search Time08:46