Coming back to coding after a multi-year break can feel paralyzing. The stack has changed, the tooling is different, the cloud is universal, and your last memory of “modern JavaScript” is jQuery and React class components. This is a 12-week plan that gets you to interview-ready without trying to learn everything at once.
Set your target first
Before you study anything, decide what role you are aiming for:
- Backend / Platform: emphasize APIs, databases, cloud, distributed systems
- Frontend: emphasize React or Vue, TypeScript, modern CSS, performance
- Full-stack: a less-deep version of both
- Data / ML: Python, SQL, ML frameworks (PyTorch / scikit-learn), data pipelines
- Mobile: Swift / Kotlin, modern app architectures
Without a target, you will study breadth and panic. With a target, your decisions become easy.
Weeks 1–2: Coding fundamentals refresh
- Pick one language you used to know well (Python, Java, JavaScript, etc.). Re-confirm fluency in syntax, idioms, and the standard library.
- Solve 30–50 LeetCode easy problems. Volume matters here, not difficulty.
- Get back to “I can implement a hash map / linked list / queue without thinking.”
Weeks 3–4: Algorithm patterns
- Two pointers, sliding window, binary search, BFS/DFS
- 30–50 medium LeetCode problems with explanation
- Resource: NeetCode 150, Sean Prashad’s LeetCode patterns, AlgoMonster
Weeks 5–6: Modern backend / cloud basics
- HTTP, REST, JSON, OpenAPI
- Postgres or MySQL — modern features (window functions, JSONB, indexes)
- AWS basics: EC2, S3, Lambda, RDS, CloudWatch (or GCP / Azure equivalents)
- Docker, Kubernetes (concepts, not depth)
- Resource: Roadmap.sh backend roadmap, AWS Solutions Architect Associate study guide
Weeks 7–8: Modern frontend (if your target is frontend or full-stack)
- Modern React (hooks, suspense, concurrent rendering)
- TypeScript
- Build tools (Vite, esbuild, Next.js)
- Resource: React docs, Kent C. Dodds Epic React, Josh Comeau’s frontend articles
Weeks 9–10: System design
- Patterns: load balancing, caching, CDN, sharding, replication, eventual consistency
- Read 4–6 case studies (Designing Data-Intensive Applications, System Design Interview Vol I/II)
- Practice mock system design with a peer or paid coach
Weeks 11–12: Behavioral and interview rehearsal
- Build the gap story (90 seconds, three parts)
- Prepare 10–15 STAR stories from your prior career
- Mock interviews — Pramp, Interviewing.io, or paid coach
- Practice talking through code, not just typing
What to skip
You will be tempted to learn everything. Resist:
- The latest LLM stack (unless you are targeting AI roles)
- Every JS framework (pick one, ignore the others)
- Esoteric algorithms (segment trees, suffix automata) — interview signal-to-noise is poor
- Frontend if you are targeting backend, and vice versa
Pace yourself
This is a marathon, not a sprint. Aim for 10–15 hours per week. Take real weekends. Burnout from cramming is the most common failure mode.
Frequently Asked Questions
How rusty is “too rusty”?
Most career-returners feel paralyzingly rusty for the first 4 weeks. By week 8 you will be remembering things you forgot you knew. By week 12 you will be ready to interview.
Should I do a bootcamp?
Usually no. Bootcamps are designed for career switchers from non-tech backgrounds. If you have prior experience, self-study is faster and cheaper.
Do I need to learn AI / LLMs to be hireable in 2026?
Not for most roles. Familiarity with using LLM coding tools (Copilot, Cursor) is now standard. Building LLM apps is a specialty, not a baseline.