Find a Site reliability engineering coach. Improve your Site reliability engineering skills.

You do not need to predict every hidden test. You need a repeatable way to find the assumptions your solution depends on and then try to break them. I use a checklist called BAD SIGNS: • B — Boundaries: First and last index, exact thresholds, inclusive versus exclusive ranges. • A — Absent or empty: Empty input, missing target, or no valid answer. • D — Duplicates: Repeated values, ties, or an input where every value is the same. • S — Single or small: One element and the smallest nontrivial case. • I — Input extremes: Zero, negatives, large integers, or unusual values allowed by the problem. • G — Giant input: Will the solution meet the time, memory, and recursion limits? • N — Natural ordering extremes: Sorted, reverse sorted, or all equal inputs. • S — Special structure or state: Cycles, disconnected components, overlaps, or source equal to destination. You do not need a test for every letter on every problem. First clarify what inputs are valid, then choose one small, high-risk example for each relevant category. The most useful step is to state an invariant: what must remain true every time your loop runs? For a sliding window with no repeated characters, for example, the window must remain duplicate-free and its left boundary must never move backward. Testing "abba" exposes implementations that violate that rule. Before submitting, do a quick line-of-code attack. For each risky operation, ask: Can this index go out of bounds? Can this map lookup miss? Can this stack be empty? Can this arithmetic overflow? Then dry-run the smallest valid input and one adversarial input, and check worst-case complexity. That is much more efficient than trying random examples until the solution “looks right.” The goal is to turn corner-case testing into a short, consistent part of how you solve every problem. I explain the method with worked coding examples in my BAD SIGNS code robustness guide. Happy coding!

The most common agentic patterns discussed in interviews are: 1. Router – Classifies requests and selects the appropriate model, tool, agent or workflow. 2. ReAct / Tool-use loop – The model selects a tool, observes the result and decides the next action. 3. Planner–Executor – A planner decomposes the objective; an executor performs and verifies each step. 4. Sequential workflow – Specialized stages run in a controlled order, such as retrieve → analyze → generate → validate. 5. Orchestrator–Workers – A coordinator delegates independent tasks to specialized or parallel workers and aggregates their results. 6. Reflection / Critic – A reviewer evaluates an output and requests bounded improvements. 7. Agentic RAG – The system determines whether retrieval is necessary, rewrites queries, searches iteratively and verifies supporting evidence. 8. Memory-augmented agent – Uses short-term state and selectively retrieved long-term memory. 9. Human-in-the-loop – Requires approval for sensitive or irreversible actions such as payments, deployments or external communication. 10. Event-driven agent – Suspends and resumes long-running work when timers, messages or external events arrive. 11. Multi-agent collaboration or debate – Multiple agents propose, challenge or validate solutions before synthesis. In a strong interview answer, do not present autonomy as the objective. Explain that production systems normally use a deterministic workflow engine, with LLM reasoning restricted to controlled steps. Add typed tools, permissions, iteration and cost limits, idempotency, retries, observability, evaluation and human approval. For a deeper architectural treatment, I recommend my white paper on SSRN: read the paper. It provides useful preparation for discussing how agentic patterns should be structured and controlled in production systems.

Sure, here we go! Senior Engineering Manager, commonly L7 Typical loop: • System design, sometimes two rounds • Coding or code review • People management • Technical leadership and execution • Behavioral/“Googliness” Expected evidence: • Leading multiple teams or a critical platform • Strong architectural judgment • Hiring, coaching, and performance management • Resolving technical disagreements • Cross-team influence • Delivering through ambiguity • Quantified technical and business results They must remain technically credible. “My engineers handle the technical details” is weak. Director, commonly L8 The loop usually emphasizes: • Technical and product strategy • Organizational design • Managing managers • Portfolio and headcount allocation • Executive and cross-functional influence • Execution across several teams • Culture, succession and organizational health Directors may still receive technical architecture interviews, but they must connect the design to team structure, investment decisions, migration strategy, and a multi-year roadmap. Typical Director questions: • Describe an organization you redesigned. • How did you handle an underperforming manager? • Tell me about a major investment you stopped. • How did you secure executive alignment and resources? • How do you prioritize when every program is critical? • Describe a technical strategy spanning several organizations. Common rejection reasons: • Senior EM behaves like a strong IC or first-line manager. • Director behaves like an experienced Senior EM. • Weak people-management evidence. • No experience managing managers. • Stories lack measurable scope and personal contribution. • Strong strategy but insufficient technical depth—or the reverse. Current Google EM postings emphasize technical leadership, architecture, product strategy, and people management. A current Director posting asks for experience managing managers, executive influence, and leadership of a large multi-team organization.

L4 engineers build out solutions for well-defined problems with code output focus. That's why L4 candidates are not assessed on system designs. For growth opportunities, L4 engineers already working at Google could slowly take on system design opportunities to prepare for an L5 promotion. The roles and expectations of engineers at each level are clearly defined in writing at Google.

1) “Tell me about a time you used data or experimentation to drive a decision in a high-ambiguity environment.” • Skills assessed: Analytical thinking, comfort with ambiguity, ability to design experiments, data-driven decision-making. • What to highlight: How you framed the problem, the data or signals you gathered, the hypotheses you tested, and the tradeoffs you considered. End with the outcome and what you learned—even if the result wasn’t perfect. 2) “How would you explain a complex AI system to a non-technical stakeholder and get buy-in?” • Skills assessed: Communication, influence without authority, user empathy, ability to simplify complexity. • What to highlight: How you broke down technical concepts into simple analogies, how you tailored the message to the stakeholder’s goals, and how you balanced transparency about risks/limitations with a clear articulation of value. 3) “Imagine the model you launched is underperforming—how do you triage and prioritize next steps?” • Skills assessed: Problem-solving, prioritization, cross-functional collaboration, resilience. • What to highlight: Your structured approach—diagnosis (data quality, model accuracy, user adoption), prioritization of fixes, how you’d partner with engineering/data science, and how you’d communicate updates to stakeholders. Emphasize learning fast and iterating.

E5s are Meta's workhorses. Therefore, they have to be very productive and own the subsystem technical direction. They need to lead E4s and E3s along with collaborating with E6s. They need to be driven to action, work on their features and support other engineers when SEVs (site events) occur. They're also expected to participate in org wide initiatives and help streamline the developer experience and operational excellence initiatives