Opendream – Backend Coding Challenge (2025 · 06)

Role Back‑End Developer (Python / C# / Go) Location Chatuchak, Bangkok Total time budget ≈ 60 – 120 min

Welcome! The coding challenges below let us observe the exact skills you'll use every day on social‑impact platforms such as Vote62 and POOPS—implementation, testing discipline, debugging, concurrency, and reasoning.

---

Repository layout

tasks/
├─ 01-run-length/
│   ├─ python/
│   ├─ go/
│   └─ csharp/
├─ 02-fix-the-bug/
│   ├─ python/
│   ├─ go/
│   └─ csharp/
├─ 03-sync-aggregator/
│   ├─ data/
│   ├─ python/
│   ├─ go/
│   └─ csharp/
├─ 04-sql-reasoning/
│   ├─ donations.db
│   ├─ expected/
│   ├─ python/
│   ├─ go/
│   └─ csharp/
└─ .github/workflows/       # ← CI (runs on any change under tasks/)

For each task pick one or more language (Python ≥ 3.10, Go ≥ 1.22, or C# / .NET 8). Edit only the stub implementation inside tasks/…/<lang>/; keep the provided tests unchanged. Any commit that touches tasks/ automatically triggers CI.

---

Challenge menu

⭐️ Any task may be discussed in follow‑up interview ⭐️

ID	Theme	Est. time	Mandatory	Who does it
01	Run‑Length Encoder	10 – 15 min	✔	Everyone
02	Fix‑the‑Bug (thread safety)	10 – 15 min	✔	Everyone
03	Sync Aggregator (concurrency & I/O)	30 – 45 min	✔*	Mid‑ & Senior
04	SQL Reasoning (analytics & optimization)	45 – 60 min	✔**	Senior only

* Juniors may stop after Task 02 ** Seniors complete all tasks

---

High-level task specifications

01 · Run‑Length Encoder (Algorithms & Data Structures)

What: Implement run‑length encoding: <char><count>

Examples:

""                              → ""
"XYZ"                           → "X1Y1Z1"
"AAAaaaBBB🦄🦄🦄🦄🦄CCCCCCCCCCCC" → "A3a3B3🦄5C12"

Key requirements:

• Case‑sensitive
• Handle multi‑digit counts
• Full Unicode support
• O(n) time complexity
• No third‑party RLE libraries

Skills tested: Basic algorithms, string manipulation, test‑driven development

See tasks/01-run-length/README.md for complete specification.

02 · Fix‑the‑Bug (Thread Safety)

What: Fix race condition in ID generator that produces duplicate values

The problem: Current implementation has a read‑increment‑write race that causes duplicates under concurrent load

Your job:

1. Do not touch the tests (they define required behavior)
2. Apply one small patch to make ID generation thread‑safe
3. Add ≤ 5 comment lines explaining why the race occurred and how your fix prevents it

Skills tested: Concurrency, race condition debugging, atomic operations

See tasks/02-fix-the-bug/README.md for complete specification.

03 · Sync Aggregator (Concurrency & I/O)

What: Process multiple text files concurrently with timeout handling and ordered results

Core challenge:

• Use worker pools (threads/goroutines, no async/await)
• Handle per‑file timeouts
• Maintain result order matching input list
• Count lines and words efficiently

Input: List of file paths (some with #sleep=N markers to simulate slow I/O) Output: JSON array with stats or timeout status per file

Skills tested: Worker pool patterns, timeout handling, result coordination, I/O processing

See tasks/03-sync-aggregator/README.md for complete specification.

04 · SQL Reasoning (Data Analytics & Index Design)

What: Write complex SQL queries and design indexes for a donations database

Tasks:

• Task A: Campaign performance analysis with percentage calculations
• Task B: 90th percentile pledge amounts (global vs Thailand)

Database: SQLite with ~100k donation records across campaigns, donors, and pledges

Skills tested: Advanced SQL (window functions, aggregations), index design, query optimization

See tasks/04-sql-reasoning/README.md for complete specification.

---

What to include in your pull request

Create one of the following:

• A SOLUTIONS.md file at repo root – or –
• A detailed PR description

For each task you complete, briefly answer:

Prompt	Why we ask
How you solved it	Outline the core algorithm / fix in 1‑2 sentences
Why this approach	Note trade‑offs (simplicity, performance, readability, etc.)
Time spent	Reality check versus our estimate

Optional extras:

• Edge cases considered
• What you'd refine with more time
• Tools/AI assistance used (if any)

Template:

## Solution notes

### Task 01 – Run‑Length Encoder
- Language: Go
- Approach: [EXPLAIN YOUR ALGORITHM]
- Why: [TRADE-OFFS CONSIDERED]
- Time spent: ~12 min
- AI tools used: [IF ANY]

### Task 02 – Fix‑the‑Bug
- Language: Python
- Approach: [EXPLAIN THE FIX]
- Why: [WHY THIS SOLUTION]
- Time spent: ~8 min

[... continue for each task completed]

---

How we evaluate

Dimension	What we look for
Correctness	Tests pass; specification satisfied exactly
Code quality	Clear naming, appropriate abstractions, readable structure
Idiomatic style	Follows language conventions and best practices
Testing discipline	Provided tests remain intact and unmodified
Problem solving	Sound reasoning in comments and approach documentation
Performance	Efficient algorithms, proper concurrency patterns (Tasks 3‑4)

Each dimension is scored 0‑3 points internally. We value quality over speed—take the time you need to demonstrate your best work.

---

Submission steps

1. Fork this repository
2. Implement solutions in tasks/…/<lang>/ (edit only implementation files)
3. Document your approach in SOLUTIONS.md or PR description
4. Commit and push your changes
5. Open a pull request against main

CI runs automatically and must pass for your submission to be considered complete.

---

Environment & FAQ

Local development

# Python
pytest tasks/01-run-length/python/
pytest tasks/02-fix-the-bug/python/
pytest tasks/03-sync-aggregator/python/
pytest tasks/04-sql-reasoning/python/

# Go
go test ./tasks/01-run-length/go
go test ./tasks/02-fix-the-bug/go
go test -race ./tasks/03-sync-aggregator/go
go test ./tasks/04-sql-reasoning/go

# C#
dotnet test ./tasks/01-run-length/csharp
dotnet test ./tasks/02-fix-the-bug/csharp
dotnet test ./tasks/03-sync-aggregator/csharp
dotnet test ./tasks/04-sql-reasoning/csharp

Dependencies

• You may use any standard library functionality
• Third‑party libraries are allowed except those that directly implement the core challenge (e.g., RLE libraries for Task 1)
• Document any external dependencies in your solution notes

AI tools & assistance

• AI coding assistants are welcome (Copilot, Claude, ChatGPT, etc.)
• We encourage using tools that help you write better code faster
• Please mention any AI assistance in your PR/solutions document
• What matters is your understanding and reasoning about the solution, please document them in SOLUTIONS.md

Best practices

• Use your best judgment and professional experience
• Quality over speed – we prefer well‑written solutions that take longer
• Test locally before submitting to ensure CI will pass
• Read the task READMEs carefully – they contain essential details

Common questions

• Q: Can I modify test files? A: No, keep all test files unchanged

• Q: What if I get stuck on a task? A: Submit what you have with clear documentation of your approach and challenges faced

• Q: Can I implement multiple languages for the same task? A: Yes, though one language per task is sufficient

• Q: How strictly are time estimates enforced? A: They're guidelines – focus on quality solutions rather than speed

---

Questions?

Open a GitHub Issue or email jobs@opendream.co.th.

We expect candidates to use their professional judgment and problem‑solving abilities. Show us how you approach real‑world development challenges!

Happy coding—we can't wait to review your solutions! 🎯

---

_{© Opendream 2025 · MIT License for this repo}