System Design Example: TikTok Recommendation System

Here’s a summary of the **infrastructure-focused ML system design** walkthrough. We are building a robust, scalable recommendation platform to support multiple product teams, focusing on infrastructure needs such as latency, reliability, and modularity.

- **We design for scalability and flexibility**, supporting 10M+ daily predictions across various surfaces (e.g., homepage, search, email), with &lt;50ms latency for real-time use cases and batch support for offline use.
- **We architect the system** with modular components: batch and streaming feature pipelines, a centralized feature store, a model training platform, a model registry, and scalable inference servers, all integrated with monitoring and alerting.
- **We implement tiered feature computation** (long-term, medium-term, real-time) using tools like Spark and Flink, and store user, item, and contextual features consistently to power accurate and fresh predictions.
- **We support multiple modeling strategies**, including two-tower models, matrix factorization, and DNNs, and use a two-stage architecture with candidate generation followed by precise ranking.
- **We deploy models reliably** using CI/CD pipelines, blue-green deployments, and shadow testing, and maintain performance through A/B testing, autoscaling servers, inference optimizations, and drift detection.

This approach ensures we build ML systems that are production-ready, highly available, and adaptable to the evolving needs of diverse teams.

If you want to learn even more from Yayun:

- Book 1 on 1 career guidance with her here: <https://calendly.com/your-friend-yayun/1on1>
- Follow her on LinkedIn: <https://www.linkedin.com/in/yayunjin/>

How You Can Supercharge Your Team - Discovering Synergies

Scope Grabs And The "Polite Fiction"

In this lesson, we discuss how leaders manage **high-performing employees who are also highly ambitious and demanding about their career progression**. Rather than viewing those ambitions as a problem, effective managers can turn them into mutually beneficial agreements—clearly defining what the employee needs to accomplish and, in return, committing to advocate for the opportunity or promotion they want.

- **Turn career demands into mutually beneficial deals:** When a high performer wants a promotion or greater opportunity, managers can establish clear expectations: deliver exceptional results on important business priorities, and leadership will help create the case for advancement.
- **Trust depends on following through:** These informal agreements only work when leaders honor their commitments. Circumstances can change, but employees can often distinguish between a manager who breaks a promise and one who genuinely loses the ability to deliver it.
- **Senior promotions often involve a queue:** At higher levels, promotions become rarer, more expensive, and sometimes constrained by organizational needs or quotas. Leaders may maintain forward-looking promotion plans, making career progression partly about understanding where someone sits in the pipeline and what they need to demonstrate to move forward.

Original podcast link:[ https://www.youtube.com/watch?v=6WaeGfLnRvc&t=24s](https://www.youtube.com/watch?v=6WaeGfLnRvc&t=24s)

ML Algorithm Example: K-means

In this segment, we walk through a high-level structure for implementing **k-means clustering from scratch**, focusing on clarity, modularity, and real-time reasoning. The emphasis is not on memorizing code, but on understanding the logic and being able to build it confidently.

- We start by defining an `__init__` method to store **core parameters** like the number of clusters (k) and max iterations.
- We **initialize centroids** by randomly selecting k data points; interviewers may also ask about smarter strategies like **k-means++**.
- In the **assign_clusters** step, we calculate distances from each point to every centroid and assign points to the nearest one.
- The **update_centroids** method recomputes each centroid as the mean of its assigned points; optionally, we may include a **loss function** to track clustering quality.
- The **fit** method loops through initialization, assignment, and updates until convergence, while **predict** assigns clusters to new data after training.
- Most importantly, we aim for **clean, modular, and well-structured code**, and narrate our logic during the interview to demonstrate deep understanding—not just syntax.

If you want to learn even more from Yayun:

- Book 1 on 1 career guidance with her here: <https://calendly.com/your-friend-yayun/1on1>
- Follow her on LinkedIn: <https://www.linkedin.com/in/yayunjin/>

Why Coding Matters In ML

In this new section on **coding fundamentals**, we explore why strong programming skills are essential for machine learning engineers. While ML involves data and models, it’s also deeply rooted in real-world software engineering.

- We clarify a common myth: **ML engineers are still engineers**, and are expected to meet the same coding standards as traditional software engineers.
- Our **software engineering background is an asset**—if applied correctly, it can set us apart from more research-focused ML candidates.
- **Coding matters in ML** because models must be integrated into production systems, maintained, and optimized for performance and scalability.
- We’re reminded to consider **algorithmic complexity**, especially when building pipelines or serving models at scale.
- The key takeaway: **don’t underestimate LeetCode-style coding prep**. Many ML interviews, especially at top tech companies, still include traditional algorithm and data structure questions. Strong coding skills are non-negotiable.

If you want to learn even more from Yayun:

- Book 1 on 1 career guidance with her here: <https://calendly.com/your-friend-yayun/1on1>
- Follow her on LinkedIn: <https://www.linkedin.com/in/yayunjin/>

Round Structure ML Algorithm

In this section, we dive into what makes the **ML algorithm coding interview** distinct from traditional coding rounds. It’s a focused, high-pressure test of both our theoretical depth and practical coding ability.

- We’ll be asked to **implement ML algorithms from scratch**—no scikit-learn, no TensorFlow—just raw code.
- This round tests both our **understanding of the algorithm’s inner workings** and our ability to **code it cleanly and efficiently**.
- With only **45–60 minutes**, we’re expected to focus on **core logic and code structure**, not perfection or full library-level implementations.
- The scope is typically **limited to common algorithms** like linear regression, logistic regression, or k-means—so it's highly trainable.
- With preparation, we can excel by knowing the **key steps, math, and edge cases** behind each algorithm and practicing how to express them in code.

If you want to learn even more from Yayun:

- Book 1 on 1 career guidance with her here: <https://calendly.com/your-friend-yayun/1on1>
- Follow her on LinkedIn: <https://www.linkedin.com/in/yayunjin/>

Become An Engineer Influencer [Part 22] - Monetization Strategies

Be Vulnerable

In this lesson, we highlight the importance of vulnerability in building a strong, supportive relationship with an engineering manager—because they can only help us if they understand what we’re going through.

- We stress that EMs can’t read minds; if we’re struggling, whether with a teammate, a project, or something personal, we should tell them so they can protect our well-being at work.
- We explain that good managers act like part-time therapists, and sharing when things are rough allows them to adjust workload, mediate conflicts, or at least offer support.
- We note that vulnerability builds trust, just like in friendships outside of work, and that trust significantly improves how effectively we collaborate.
- We clarify that we shouldn’t overshare, but we also shouldn’t hide everything—sharing selectively helps EMs understand who we are and how to help.
- We reinforce that showing vulnerability increases the “trust meter,” strengthens the partnership, and ultimately unlocks more powerful support from our EM.

[Paper Reading] Netflix : Query Facet Mapping and its Applications in Streaming Services

Led by [Dharin Shah](https://www.linkedin.com/in/dharinshah/)

For a Taro Premium discount from Dharin, use his referral link here: <https://www.jointaro.com/r/dharins694/>

This deep-dive goes through how Netflix built their internal query to facet mapping layer to have better understand intent of the query. For example - matching 'anim' to 'Demon Slayer' as higher rank than perhaps 'animal planet'. In this session we dive into the white paper released by Netflix around their implementation.

Link to Paper: [**https://dl.acm.org/doi/pdf/10.1145/3477495.3536330**](https://dl.acm.org/doi/pdf/10.1145/3477495.3536330)

Become An Engineer Influencer [Part 16] - Editing

Mock Interviews Are Non-Negotiable

To wrap up our preparation strategy, we dive into how to *structure* our prep effectively—because randomly studying “a bit of everything” doesn’t work.

- **Mock interviews are non-negotiable**: They’re the fastest way to sharpen your skills, build confidence, and improve how you communicate under pressure.
- **Avoid practicing in real interviews**: We should use mocks to make mistakes and learn, not high-stakes opportunities.
- **Find partners in your network or online**: Leverage communities or platforms like ADP List to schedule mock sessions with others.
- **Record your mocks**: Watching ourselves afterward reveals blind spots we may not notice in the moment—this step alone can lead to major improvements.

If you want to learn even more from Yayun:

- Book 1 on 1 career guidance with her here: <https://calendly.com/your-friend-yayun/1on1>
- Follow her on LinkedIn: <https://www.linkedin.com/in/yayunjin/>

Amazon Behavioral Interview [Part 24] - Sacrificing Short-Term For Long-Term (Evaluation)

This video segment provides feedback on an individual's response to a behavioral interview question about a time they sacrificed short-term gains for long-term value.

- **Strengths:** The interviewee effectively described the initial situation, highlighting the inefficiencies of the existing regional team structure and the need for a more globally aligned approach.

- **Areas for Improvement:**

  - **Focus on Long-Term Impact:** The interviewee could have more clearly and quantitatively demonstrated the long-term positive impact of the changes, such as specific improvements in key performance indicators (KPIs) beyond just reduced resolution times.
  - **Minimize Negative Framing:** The interviewee could have reframed their description of the challenges faced, minimizing the emphasis on negative aspects like "disgruntled people" and "pain points."
  - **Demonstrate Stronger Leadership:** While the interviewee mentioned gathering feedback and collaborating with stakeholders, they could have further emphasized their leadership skills, such as:
    - Proactively communicating the vision and rationale for the change to the team.
    - Effectively managing stakeholder expectations and addressing concerns.
    - Motivating and empowering the team throughout the transition process.

- **Overall:** While the interviewee effectively described the situation and the challenges faced, the response could be strengthened by:

  - **Focusing on positive outcomes:** Emphasizing the significant improvements in efficiency, customer satisfaction, and team performance.
  - **Demonstrating stronger leadership and communication skills:** Highlighting proactive measures taken to address team concerns and ensure successful implementation.
  - **Minimizing negative framing:** Presenting the challenges in a more constructive and solution-oriented manner.

Learn more about Alan's job search program, Kadima Careers, using this [discount link](https://signup.kadimacareers.com/mQBsDvu).

Amazon Behavioral Interview [Part 16] - Going Above and Beyond For A Customer (Evaluation)

This video segment provides feedback on an individual's response to a behavioral interview question: "Tell me about a time when you went above and beyond for a customer."

- **Strengths:**

  - **Impactful Story:** The interviewee described a compelling situation where they went above and beyond to help a customer who was facing a significant issue with their credit card insurance.
  - **Demonstrated Initiative:** They proactively investigated the issue, identified a potential resolution, and engaged with key stakeholders, including a senior executive, to advocate for the customer.
  - **Highlighted Positive Outcomes:** The interviewee emphasized the positive outcomes of their actions, including:
    - Resolving the customer's issue and restoring their satisfaction.
    - Retaining a valuable customer who was considering canceling their card.
    - Driving a policy change to improve the customer experience for future cardholders.

- **Areas for Improvement:**

  - **Further Emphasize Leadership and Influence:** While the interviewee mentioned engaging with the SVP, they could have further emphasized their leadership and influence by describing how they:
    - Presented a compelling case for the customer's situation.
    - Overcame potential objections from the SVP.
    - Effectively communicated the importance of this issue to senior leadership.
  - **Quantify Impact Further:** While the interviewee mentioned the customer's annual spending, they could have further quantified the impact of their actions by:
    - Estimating the potential lifetime value of the customer.
    - Describing the positive impact on customer satisfaction scores.
    - Highlighting the potential long-term benefits of the policy change.

- **Overall:** The interviewee provided a strong foundation for their response. By incorporating the suggested improvements, they can further enhance their ability to effectively communicate their accomplishments and demonstrate their leadership, problem-solving, and customer-centric approach in a job interview setting.

Learn more about Alan's job search program, Kadima Careers, using this [discount link](https://signup.kadimacareers.com/mQBsDvu).

Management vs IC Transitions

In this lesson, Philip reflects on his experience moving between individual contributor (IC) and management roles throughout his career. He explains why he initially pursued management, why he ultimately prefers being an IC, and how he has managed to switch between the two tracks multiple times—even when doing so meant willingly stepping down in level.

- **Choose management because you want to lead—not simply because it feels like the next promotion.** Before becoming a manager, consider whether your team would genuinely want you to lead them and whether you would want to work for yourself.
- **Be careful of turning management into a one-way door.** Moving away from hands-on technical work can make returning to the IC track difficult, particularly if maintaining your title, level, or compensation becomes more important than making the right career move.
- **IC and management skills can converge at senior levels.** Senior ICs often need many of the same capabilities as strong leaders—including communication, vision, coaching, business understanding, and the ability to lead complex initiatives—although highly specialized technical roles may follow a different path.

Original podcast link: [**https://www.youtube.com/watch?v=v2JxdjTi_1I**](https://www.youtube.com/watch?v=v2JxdjTi_1I)

Project Presentation

Some companies include a **project presentation round** in their ML interview process, where we’re expected to walk interviewers through a real project from start to finish. This is a powerful opportunity to showcase our end-to-end technical skills, communication ability, and strategic thinking.

- **Use visuals** to anchor the conversation—simple diagrams in Excalidraw or Google Docs help explain architecture clearly and keep interviewers engaged.
- **Explain concepts like you’re teaching a junior engineer**, avoiding jargon and offering quick context before diving into complex ideas (e.g., why negative sampling is needed).
- **Stay composed during interruptions**, as these are often tests of our communication clarity rather than signs of concern.
- **Show real trade-off thinking**—don’t just name-drop methods; explain *why* one approach was selected over others based on constraints like latency or scalability.

This round is about more than technical depth—it’s about demonstrating that we can **lead projects, make informed decisions, and communicate them with confidence and clarity**.

If you want to learn even more from Yayun:

- Book 1 on 1 career guidance with her here: <https://calendly.com/your-friend-yayun/1on1>
- Follow her on LinkedIn: <https://www.linkedin.com/in/yayunjin/>

Model Training

In this segment, we cover **model training**, a phase where many candidates fall into autopilot. To stand out, we need to show thoughtfulness, adaptability, and a clear connection between our choices and the problem at hand.

- **Mediocre performance** involves picking a default model (e.g., logistic regression or random forest), using out-of-the-box settings with no tuning or justification.
- **Strong candidates** choose models deliberately, explaining how their selection fits the **data characteristics, business goals, or system constraints**.
- We improve performance through **hyperparameter tuning**, using techniques like grid search or informed manual adjustments based on EDA insights.
- When appropriate, we apply **ensemble methods** (e.g., boosting, bagging, stacking) to push performance further.
- Most importantly, we demonstrate that our modeling choices are **intentional and grounded in context**, not just rote or random experimentation.

If you want to learn even more from Yayun:

- Book 1 on 1 career guidance with her here: <https://calendly.com/your-friend-yayun/1on1>
- Follow her on LinkedIn: <https://www.linkedin.com/in/yayunjin/>

Don't Freak Out

We emphasize the mindset shift that matters most when facing the reality of layoffs and uncertainty in tech.

- The most important first step is not to freak out, even though layoffs can feel shocking and destabilizing, especially for people who assumed big tech jobs were inherently safe.
- Acceptance is key: layoffs are a normal part of professional life, often driven by factors far outside an individual contributor’s control, including executive decisions and market conditions.
- Obsessing over layoffs or drastically changing your behavior can actually backfire, while staying steady and focused on your work is usually the healthier and more effective approach.
- If a layoff is going to happen, the decision is typically already made, so recognizing what you cannot control can be freeing and allows you to focus your energy on what you actually can control.

Maximize Your Equity And Benefits In Tech [Part 15] - Alternative Minimum Tax Basics

Alternative Minimum Tax (AMT) can create surprise tax liabilities, especially when exercising incentive stock options. Here's the core points from the lesson:

- AMT is a parallel tax system with fewer deductions, aimed at ensuring high earners pay baseline taxes, and it most often affects those exercising ISOs below market value
- Unlike the progressive regular tax system, AMT uses flat rates (26% or 28%) and doesn’t allow deductions like standard or state/local taxes, leading to potentially higher taxes
- The difference between ISO strike price and market value counts as AMT income, which may trigger a large tax bill even if the shares aren't sold
- Any excess AMT paid over regular tax becomes a credit that can offset future tax liabilities, but recovering it depends on your future income and tax situation

[Check out my AMT Calculator here](https://adastrafp.com/amt-calculator/) (for informational purposes only, not tax advice.)

Next steps with Andy:

- [Connect with Andy on LinkedIn](https://www.linkedin.com/in/andymoranio)
- [Leave feedback and/or sign up for Andy’s newsletter](https://subscribepage.io/taroequitycourse)
- [Schedule a consultation](http://adastrafp.com/contact-us/)

Notebook Interview Preparation Tips

In this final segment, we shift from execution to preparation—learning how to practice effectively for notebook-style ML interviews. The focus is on developing real-world fluency, structured habits, and clear communication.

- We should **practice with real, messy datasets** from platforms like Kaggle or UCI to simulate the open-ended nature of actual interview problems.
- We’re encouraged to build a **reusable workflow template**—a consistent structure (e.g., data loading, EDA, feature engineering, modeling, evaluation)—to streamline our process and reduce cognitive load.
- We must get used to **explaining our decisions out loud**, as live interviews often involve narrating our thought process clearly and confidently.
- Throughout, we should **document our work as we go**, making it easier for interviewers to follow our logic and showing that we value clarity.
- Most importantly, we’re reminded to **think beyond accuracy**—considering the user impact, business value, and maintainability of our solution.
- A bonus tip: **practice talking while coding** to improve our ability to communicate effectively under real-time conditions. This small habit can make a big difference in live interviews.

If you want to learn even more from Yayun:

- Book 1 on 1 career guidance with her here: <https://calendly.com/your-friend-yayun/1on1>
- Follow her on LinkedIn: <https://www.linkedin.com/in/yayunjin/>

Evaluation

In this section, we explore **model evaluation**, a phase that’s often underestimated but critically important. This is our chance to show we’re not just optimizing metrics—we’re aligning our work with real-world impact.

- **Mediocre evaluation** involves simply calling `.score()` or reporting accuracy, with little explanation or context—a major red flag in interviews.
- **Stellar candidates choose metrics that fit the problem**, such as F1 score or AUC-ROC for imbalanced classification, or MAE over RMSE for certain regression tasks.
- We demonstrate awareness of **trade-offs**, like how boosting recall might reduce precision, and explain the implications clearly.
- Most importantly, we **tie evaluation to business goals**, showing we understand the cost of errors and what metrics matter most in practice.
- This thoughtful approach proves that we’re not just building models—we’re optimizing for meaningful outcomes.

If you want to learn even more from Yayun:

- Book 1 on 1 career guidance with her here: <https://calendly.com/your-friend-yayun/1on1>
- Follow her on LinkedIn: <https://www.linkedin.com/in/yayunjin/>

Tech Career Growth Videos

Our instructional videos provide software engineers with insights needed for career growth. Masterclasses go incredibly in-depth on a specific tech topic while case studies are real examples from established engineering leaders.