Preparing for an IT technical test in Switzerland

Types of technical tests

1. Take-home assignment:

• 2-8 hour project to do at home
• Deadline: 3-7 days
• Topic: web app, API, data transformation
• Criteria: code structure, architectural choices, tests, README

2. Live algorithm:

• 45-60 min via CoderPad, HackerRank, CodeSignal
• LeetCode-type problems (Easy to Hard)
• Criteria: think-aloud reasoning, Big-O complexity, edge cases

3. System design:

• 45-60 min discussion
• Topic: "How would you build Twitter / Uber / Netflix?"
• Criteria: scaling capacity, justified technical choices, trade-offs

4. Pair programming:

• 60-90 min side-by-side (virtual) with a senior dev
• Topic: feature to add to existing code, bug to fix
• Criteria: code understanding, communication, collaboration

5. Code review:

• 30-45 min, analysis of PR with intentional issues
• Criteria: bug detection, feedback quality, sense of detail

Preparation by test type

For algorithms:

• LeetCode: aim for 200-300 solved problems (Easy/Medium dominant)
• NeetCode 150: curated list of classic problems
• Cracking the Coding Interview (Gayle McDowell): reference book
• Daily practice: 30-45 min, 5-6 days/week for 8-12 weeks

For system design:

• System Design Interview (Alex Xu): 2 essential volumes
• High Scalability (highscalability.com): real case studies
• Designing Data-Intensive Applications (Kleppmann): in-depth reference
• YouTube: Gaurav Sen, ByteByteGo channels

For take-home:

• GitHub: 3-5 clean public projects before interview
• Unit tests: show your rigour
• Detailed README: architectural decisions, how to run, known limitations
• Polish commit history: atomic commits, explicit messages

Test day

Live algorithm:

1. Clarify the problem (5 min): examples, constraints, edge cases
2. Think aloud: your reasoning counts as much as the solution
3. Brute force first: show you understand
4. Optimise: Big-O, memory space, justification
5. Code cleanly: clear variable names, modular functions
6. Mentally test: trace execution on examples + edge cases

System design:

1. Clarify requirements: functional and non-functional
2. Estimate load: RPS, storage, bandwidth
3. High-level diagram: components, data flows
4. Deep-dive on 1-2 critical components
5. Bottlenecks and scaling: where it breaks, how to scale
6. Trade-offs: justify your choices

Recommended tools

• LeetCode Premium (USD 35/month): company-targeted questions
• AlgoExpert: detailed video explanations
• GitHub Copilot: familiarise yourself (used in interviews at some)
• Excalidraw, Whimsical: clean system design diagrams
• CoderPad: practice the real interview environment

Swiss specifics

• Banks (UBS, Raiffeisen): rigorous tests, focus algo + system design
• Multinationals (Google, Microsoft Zurich): classic Anglo-Saxon process, 5-6 rounds
• Start-ups (Logitech, Sonova): take-home + pair programming dominant
• Pharma (Roche, Novartis): focus on data engineering, ETL, SQL
• Crypto (Zug): required Web3 knowledge, blockchain basics

Common pitfalls

• Coding without thinking: 2 min of planning saves 20 min of debug
• Not asking for help: "Can I get a hint?" is better than stuck 15 min
• Over-engineering: don't create 5 abstractions for a fizz-buzz
• Neglecting tests: show you think about them even if not writing all
• Lack of communication: an evaluator who doesn't understand your reasoning will rate you poorly