Skip to content

Interview Guidelines

How to turn the data in this registry into a preparation plan for Tanzanian technical interviews.

1. How to Read a stack.json

A stack is a signal of what the company values. Read it in this order:

  1. backend — dictates algorithmic and design-pattern style questions. A Go shop and a Java shop will interview very differently.
  2. database — predicts the depth of data-modeling and SQL questions. PostgreSQL-heavy companies ask about indexing and transactions; MongoDB-heavy ones ask about document design and consistency.
  3. cloud_infra — signals system design scope. AWS + Kubernetes means they expect you to reason about deployment, scaling, and failure modes.
  4. mobile_tech — if populated, expect mobile-specific questions (offline sync, low-bandwidth UX, state management).
  5. frontend — the lightest signal for backend roles, heaviest for fullstack roles.

2. The Two Axes of a Tanzanian Interview

Most interviews here split cleanly along two axes. Prepare for both.

A. System Design

Focus on problems the company actually faces. For Tanzanian companies this usually means:

  • Unreliable networks. How does your system behave on 3G or intermittent connectivity?
  • Mobile money integration. Callbacks, idempotency, and reconciliation with M-Pesa, Tigo Pesa, Airtel Money, and Halopesa.
  • USSD. Stateful session flows over a stateless protocol.
  • Multi-currency. TZS, USD, KES cross-border flows (especially for fintech).
  • Concurrency and transactional integrity. Double-spend prevention, atomic balance updates.

Typical prompts: - "Design a wallet service that supports top-ups from mobile money." - "Design an inventory system for a chain of shops with unreliable internet." - "Design a ride-hailing dispatch system for Dar es Salaam."

B. Practical Coding

These are hands-on exercises, often timed. Expect:

  • Consume a public REST API and transform the response.
  • Implement a small CRUD service with a database.
  • Write SQL for a multi-table reporting query.
  • Build a component that handles loading, error, and offline states.

Practice against the stack listed in stack.json — if the company uses Django, do the exercise in Django, not Flask.

3. A Seven-Day Preparation Loop

  1. Day 1 — Read the company's stack.json and interview-questions.md. Find one engineering blog post or conference talk from the company.
  2. Day 2 — Review fundamentals for their primary backend language and database.
  3. Day 3 — Build a small project using two of their core technologies.
  4. Day 4 — Solve three system-design prompts from this repo aloud, recording yourself.
  5. Day 5 — Do two timed coding challenges in their stack.
  6. Day 6 — Prepare questions for the interviewer, grounded in something specific you saw on their blog or GitHub.
  7. Day 7 — Rest. Reviewing at low energy reinforces mistakes.

4. Interview-Day Tactics

  • Narrate your thinking. Silence reads as "stuck" even when you're making progress.
  • State assumptions. "I'm assuming users can retry a failed payment — is that correct?" Prevents you from solving the wrong problem.
  • Write trade-offs, not verdicts. "Postgres gives us transactions; Mongo gives us flexible schemas. For a ledger, I'd pick Postgres because X." Beats "Postgres is better."
  • For mobile money questions, always mention idempotency keys and reconciliation jobs. These are the two things that separate juniors from seniors in Tanzanian fintech interviews.

5. What This Repo Does Not Replace

  • Reading the official docs for the technologies listed.
  • Building real projects.
  • Talking to engineers who work at the company.

Use the registry as a map, not as the territory.