This roadmap is for CS students in their 1st–3rd year who want to graduate with a strong AI+engineering profile. The goal is not to learn everything — it is to learn the things that compound.
12 months
Year 1: Foundations That Don't Expire
- CS fundamentals: data structures and algorithms — enough for coding rounds, not to the exclusion of everything else.
- One programming language mastered: Python or JavaScript, not a shallow tour of five.
- Version control and a building habit: commit often, push to GitHub, ship small things weekly.
- First side project: small, finished, and yours — proof you can take an idea to working code.
12 months
Year 2: AI + Full-Stack
- RAG basics: chunking, embeddings, retrieval — enough to build a working document Q&A.
- LLM API usage: prompting, streaming, cost, integrating a model into a real app.
- One full-stack project: front end + back end + deployed to the internet.
- DSA practice: consistent and steady, not frantic cramming.
- Contribute to open source or build in public: visible work that compounds your reputation.
Lessons for this stage
- What is RAG and When Does It Beat Pure Prompting15 min
- Embeddings Explained: From Text to Vectors15 min
- Build a RAG Pipeline from Scratch1h
- Prompt Engineering Patterns: System Prompts & Chat Templates15 min
- Next.js App Router Mental Model: Server vs Client Components25 min
- Deploying to Vercel: Env Vars, Database, Preview Branches25 min
12 months
Year 3: Specialize + Ship
- Pick a direction: AI engineering or full-stack — go deep on one.
- Build one substantial portfolio project: the centerpiece of your applications.
- Placement and interview prep: company-targeted, not generic grinding.
- Resume and LinkedIn polished: clear, project-led, and recruiter-readable.
Lessons for this stage
3–6 months
Final Semester: Placement Mode
- Mock interviews: coding and behavioral, with real feedback.
- Company research: know what each target actually hires for.
- Competitive programming only if required by your target companies — otherwise skip it.
- Referrals and networking: the highest-leverage activity in the final stretch.
Skill priorities
Learn now
- DSA fundamentals
- One programming language fluently
- Git and public code
- Building side projects
Learn soon
- RAG and LLM APIs
- Full-stack web (front + back + deploy)
- System design basics
Learn later
- Agents and multi-agent systems
- Fine-tuning
- Cloud certifications
Ignore for now
- Competitive programming as primary strategy unless targeting those companies
- Tutorial certificates
- Learning every new framework
- Social media performance instead of actual building