Arun Baby

Becoming Unlabelable

Overview

A structured 60-day learning journey through three interconnected domains of computer science and machine learning:

  1. Data Structures & Algorithms - Core computer science fundamentals
  2. ML System Design - Production-scale system architecture
  3. Speech Technology - Audio and speech ML from research to production

Progress Tracker

Posts Completed: 36 / 180 (20.0%)

Last Updated: October 20, 2025

Week 1: Foundations & Real-Time Systems (Posts 1-7)

Week 2: Data Structures & Validation (Posts 8-14)

Week 3-9: Coming Soon

Topic mapping in progress…

Three Parallel Tracks

🔢 Data Structures & Algorithms

Goal: Master fundamental algorithms and data structures for coding interviews

Focus Areas:

  • Arrays, Hash Tables, Strings
  • Linked Lists, Stacks, Queues
  • Trees, Graphs, Tries
  • Dynamic Programming
  • Sorting, Searching
  • Greedy Algorithms

Approach:

  • LeetCode-style problems
  • Multiple solutions (brute force → optimal)
  • Time/space complexity analysis
  • L6/L7 specific insights
  • Production considerations

Browse all DSA problems →

🏗️ ML System Design

Goal: Design scalable, production-grade ML systems

Focus Areas:

  • Recommendation Systems
  • Search & Ranking
  • Computer Vision Systems
  • NLP Systems
  • Real-Time ML
  • Feature Engineering
  • Model Serving
  • MLOps & Monitoring

Approach:

  • End-to-end architecture
  • Requirements gathering
  • Component deep-dives
  • Scaling strategies
  • Trade-off analysis
  • Failure mode handling

Browse all ML System Designs →

🎤 Speech Technology

Goal: Deep technical expertise in speech and audio ML

Focus Areas:

  • Automatic Speech Recognition (ASR)
  • Text-to-Speech (TTS)
  • Speaker Recognition & Diarization
  • Voice Activity Detection (VAD)
  • Speech Enhancement & Denoising
  • Conversational AI
  • Real-Time Streaming
  • On-Device Optimization

Approach:

  • Production architectures
  • Model selection and trade-offs
  • Latency optimization
  • Streaming considerations
  • Edge deployment
  • Scaling strategies

Browse all Speech Tech posts →

Why This Matters

Learning Philosophy

This structured approach emphasizes:

  • Fundamentals First: Strong foundation in algorithms and data structures
  • System Thinking: Understanding how concepts scale in production
  • Depth & Breadth: Deep expertise in speech tech, broad ML knowledge
  • Practical Application: Real-world engineering trade-offs

Thematic Connections

Each day’s problems are thematically linked to reinforce concepts:

Example - Day 1:

  • DSA (Two Sum): Hash table for O(1) lookup → Feature stores
  • ML System (Recommendations): Fast embedding lookup, caching strategies
  • Speech (ASR): Low-latency inference, state management

This approach builds systems thinking - connecting theory to practice.

How to Use This Resource

For Learners

  1. Sequential Learning: Follow the 60-day progression for structured growth
  2. Pick Your Track: Focus on DSA, ML Systems, or Speech based on interests
  3. Practice Actively: Code solutions, design systems, implement concepts
  4. Connect Ideas: Notice thematic links across domains

For Practitioners

  • Reference material for system design decisions
  • Production engineering patterns and trade-offs
  • Code examples and architecture templates
  • Research foundations for speech/audio ML

For Interview Preparation

  • Comprehensive coverage of common patterns
  • Multiple difficulty levels (easy → hard)
  • Discussion points and trade-off analysis
  • Real-world system examples

Methodology

Daily Structure

Time Commitment: 2-3 hours/day

  1. DSA Problem (45-60 min):
    • Understand problem
    • Brute force solution
    • Optimize
    • Code + test
    • Write analysis
  2. ML System Design (60-90 min):
    • Requirements gathering
    • Architecture design
    • Component details
    • Write comprehensive post
  3. Speech Tech (45-60 min):
    • Research current approaches
    • Architecture analysis
    • Code examples
    • Write technical deep-dive

Quality Standards

  • DSA: Multiple approaches, full complexity analysis
  • ML Systems: Production-ready architectures, scalability analysis
  • Speech: State-of-the-art techniques, real-world trade-offs

Inspiration

  • LeetCode for DSA problems
  • System Design Interview books (Alex Xu, Grokking)
  • Real-world systems (Google, Meta, Amazon papers)
  • Research papers for speech tech

Topics Roadmap

Thematic Weeks

Week 1-2: Real-Time Systems & Graphs

  • Graph algorithms ↔ Recommendation graphs ↔ Streaming ASR

Week 3-4: Sequence Problems & NLP

  • Dynamic programming ↔ Seq2seq models ↔ TTS/ASR

Week 5-6: Trees & Hierarchical Systems

  • Tree algorithms ↔ Model serving routing ↔ Speaker clustering

Week 7-8: Optimization & Performance

  • Greedy algorithms ↔ Feature engineering ↔ Model optimization

Week 9: Advanced & Integration

  • Advanced algorithms ↔ Distributed training ↔ Production deployment

Resources

Books

  • Cracking the Coding Interview (Gayle Laakmann McDowell)
  • System Design Interview Vol 1 & 2 (Alex Xu)
  • Designing Machine Learning Systems (Chip Huyen)
  • Machine Learning Engineering (Andriy Burkov)

Online Platforms

Papers & Talks

  • Google Research
  • Meta AI Research
  • ArXiv (Speech & Audio)
  • Conference talks (NeurIPS, ICASSP, Interspeech)

Connect

I’m documenting this journey publicly to:

  • Hold myself accountable
  • Help others preparing for similar roles
  • Demonstrate technical depth and breadth
  • Build a portfolio of technical writing

Follow along:

Questions or suggestions? Contact me

Changelog

October 2025

  • 🚀 Started 60-day learning challenge
  • ✅ Completed first week: Foundations
  • 📝 Set up three-track structure

Let’s build something great together. Day by day. 🚀

Content created with the assistance of large language models and reviewed for technical accuracy.