Búsqueda de Empleo & Cambio de Carrera — Kit IA
Diez picks para quien cambia de trabajo, sector o entra a ingeniería IA. Constructor de CV + base de conocimiento de LinkedIn + sistema de skill de búsqueda + tutor de algoritmos + entrevistas simuladas + tutor conductual + plantillas de carta + automatización de outreach + coach de comunicación + memoria persistente. En orden.
What's in this pack
This is the rig you'd actually build if you were leaving a job in two weeks and wanted to land somewhere better — not a directory of every job-search SaaS. Ten picks, organized into five layers:
- Resume layer — Reactive Resume (open-source, AI-assisted, ATS-friendly export)
- Brand / LinkedIn / knowledge base — Career-Ops (the spine), Notion MCP (job tracker), Mem0 (persistent context across applications)
- Algorithm and technical prep — DeepTutor (Socratic algorithm tutor), Microsoft Study Mode (Claude / GPT in active-recall mode)
- Mock interview and behavioral — Mr. Ranedeer (personalized tutor prompt for behavioral / STAR), Communication Excellence Coach (interview delivery)
- Cover letter and cold outreach — Fabric (100+ writing patterns including cover letter / pitch), Sales Automator (cold-email sequencing for networking)
The whole stack is open-source or free-tier, runs locally where possible, and works in any agent (Claude Code, Codex, Cursor, Gemini CLI). No platform lock-in. No paid "career coach" SaaS that disappears the day you sign an offer.
Install in this order
Order matters more here than in most packs. Each layer feeds the next.
- Career-Ops (id 3817) — the spine. This is a Claude Code skill that orchestrates your whole job search: tracks applications, drafts replies, summarizes JDs. Install first so every other tool drops into a real pipeline instead of a Notion graveyard.
- Reactive Resume (id 866) — the resume. 36k★ open-source, MIT, AI integration with Claude / GPT / Gemini, exports a clean ATS-friendly PDF. Build the master resume once; tailor per role.
- Notion MCP (id 838) — the job tracker. Connect Career-Ops to a Notion workspace so every application, JD, contact, and round becomes queryable by your agent. Replaces the 47-tab spreadsheet.
- Mem0 (id 703) — persistent memory. Now your agent remembers "I told Acme Corp's recruiter I'd follow up Friday" across sessions and devices. Critical once you're juggling 15+ active threads.
- DeepTutor (id 4178) — algorithm prep. Doesn't give you LeetCode answers — walks you through the approach (constraints → patterns → complexity → code). Use 30 min/day, not 4 hours on weekends.
- Microsoft Study Mode (id 4434) — active-recall drilling. Flips into Socratic mode for system-design and behavioral. Pair with DeepTutor: one for algorithms, one for everything else.
- Mr. Ranedeer (id 24) — personalized behavioral interview tutor prompt. Drop it into Claude / GPT and it teaches STAR / SCQA framing tuned to your background.
- Communication Excellence Coach (id 4295) — delivery polish. Rehearse your answers, get feedback on filler words, hedging, structure. The last 10% that decides offer vs. rejection.
- Fabric (id 312) — writing patterns. Run
fabric -p write_pull_requeststyle patterns for cover letters, follow-ups, recruiter outreach. 100+ patterns means you stop asking ChatGPT "write me a cover letter" cold. - Sales Automator (id 4307) — cold outreach sequencing. Built for sales but the email patterns map 1:1 to networking outreach for career pivots — first message, follow-up at day 4, value-add at day 9.
How they fit together
Career-Ops (skill spine)
│
┌──────────┼──────────────┬─────────────────┐
│ │ │ │
Reactive Notion MCP + DeepTutor + Fabric +
Resume Mem0 (memory) MS Study Mode Sales Automator
│ │ │ │
│ LinkedIn / LeetCode / Cover letter,
│ JD tracking system design cold email,
│ recruiter ping
│
└─→ PDF + ATS export → apply → Communication Coach → interview
↑
Mr. Ranedeer (behavioral / STAR)
The top of funnel — positioning, resume, LinkedIn — is one thing. The middle — algorithm + system design + behavioral practice — is a second. The bottom — cold email, networking, interview delivery — is a third. This pack assigns one to three tools per layer so each layer has a clear default.
Tradeoffs you'll hit
- AI resume vs. ATS reality — every "AI resume builder" promises to beat ATS. The truth is most ATS systems do simple keyword + section parsing. Reactive Resume's plain-text export wins because it doesn't fight the parser. Avoid templates with multi-column layouts, icons in headers, or images of skills bars — they break parsing.
- Mock interview AI vs. real humans — DeepTutor + MS Study Mode are infinite, free, and never judge you. They cannot replicate the cognitive load of being grilled by a hiring manager at 9pm on a Tuesday. Use AI for reps (50 problems / 20 mock behaviorals), then book 2-3 real mocks before the actual loop.
- Algorithm depth vs. breadth — for one specific company you can drill their top-50 tagged questions. For a 3-month pivot you need the breadth: arrays / strings / trees / graphs / DP / BFS / DFS / two pointers / sliding window. Pick a curriculum (Neetcode 150 or Blind 75), then use DeepTutor on the questions you get stuck on.
- Career-Ops skill vs. SaaS career platforms — Career-Ops lives in your editor and integrates with Notion MCP + Mem0. The SaaS alternatives (Careerflow, Teal, Simplify) are polished but you pay monthly and walk away with nothing. Career-Ops gives you data ownership and agent extensibility.
Common pitfalls
- AI-written resume reads generic and forgettable — every recruiter has read 200 AI-generated resumes this month. The fix isn't "better prompts" — it's specifics: dollar amounts, percentages, named systems, before/after metrics. Use AI to structure and tighten the bullets you already wrote, not to invent them.
- LinkedIn keyword-stuffing tanks profile views — LinkedIn's algorithm has gotten good at detecting paragraphs of stuffed skills. Write the headline and About like a human. Put keywords in experience bullets where they belong.
- Cold-emailing without researching the company — 30 seconds of research on the company's last 3 product launches lifts reply rate dramatically. Don't skip this for speed. One 80-word email per company beats 50 templated emails.
- Mass-applying with a generic cover letter — most are skim-read in 8 seconds. If you're using Fabric for cover letters, customize at least the first sentence and one mid-paragraph reference to the role / product. The rest can be templated.
- Not tracking what's working — without Notion MCP + Mem0, by week 6 you won't remember which subject line got replies, which referral source converted, which role types you actually interview well for. Track from day one.
10 recursos listos para instalar
Preguntas frecuentes
Will an AI-written resume get rejected by ATS systems?
Not because it's AI-written — ATS systems don't detect AI. They reject resumes that fail keyword matching or break parsing. The risk with AI resumes is that they read generic and forgettable to the human reviewer who reads it after ATS passes. Use Reactive Resume's clean text-based export for ATS compatibility, then make sure each bullet has specifics (numbers, system names, scope) that a human will remember.
What LinkedIn metrics should I actually optimize for?
Profile views per week (target 50+ if you're actively searching), search appearances (proxy for keyword fit), and InMail / connection-request reply rate (proxy for how compelling your profile is once someone lands). Vanity metrics like followers and post likes matter only if you're job-hunting in public. Optimize headline + About + featured first; those drive 80% of profile conversion.
Which mock interview tool should I actually pick?
For algorithms: DeepTutor for the Socratic walkthrough (it asks you what approach you'd try before showing code), or any IDE with Claude / GPT in the chat. For behavioral: Mr. Ranedeer with a STAR-framing prompt — it role-plays the interviewer and gives you delivery feedback. For final polish 1-2 weeks before the loop, book 2-3 sessions with real humans (peers in your target field beat paid coaching for technical realism). AI for reps, humans for the loop.
How long does a career pivot to AI engineering realistically take?
Depends on starting point. If you're a working software engineer adding LLM / agent skills: 8-12 weeks of focused side projects + 4-6 weeks of active applying. If you're transitioning from a non-engineering role: assume 6-12 months including foundational coding + one shipped portfolio project + targeted networking. The pack helps with the application loop, not with closing the underlying skill gap — be honest about which one you're solving for.
What's a realistic reply rate for cold-email outreach?
For warm outreach (someone you have a 2nd-degree connection to, with a specific reason): 20-40% reply rate is normal. For cold outreach (no connection, generic subject): 2-8%. The difference is research, specificity, and value — not volume. A short, well-researched email referencing the recipient's recent work or a specific company initiative beats 100 templated messages. Track reply rate by template in Notion; iterate weekly.
12 packs · 80+ recursos seleccionados
Explora todos los packs curados en la página principal
Volver a todos los packs