Reddit·Machine Learning Engineer·Onsite - Product Sense / Strategy·Senior

Senior

Jun 2026

Summary

Reddit MLE interview that threw a product sense curveball. The main question was a full onboarding teardown for new Reddit users, which felt odd for an ML role but apparently they do this.

Questions Asked (1)

How would you improve Reddit's onboarding experience for new users? Cover who these users are and what they want, where the current flow breaks down, two or three concrete improvements you'd prioritize, and how you'd measure success and run experiments.

Product Sense & IdeationA/B Testing & ExperimentationProduct Analytics & Metrics

Author's notes

Not what I expected from an ML round.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by segmenting new users (e.g., lurkers, content seekers, community joiners) and articulating their goals and pain points. Then map the current onboarding flow to identify drop-off points, propose 2-3 ML-driven improvements, and define success metrics and an experimentation plan. Frame your answer around user value and measurable impact.

Pro tip: Tie each improvement to a specific ML technique (e.g., embeddings for interest matching, bandits for exploration) and emphasize how you'd validate offline before A/B testing. Show you understand Reddit's unique community dynamics and content diversity.

1. Define user segments and goals

Identify distinct new user groups (e.g., lurkers, posters, niche community seekers) and their motivations. Clarify what success looks like for each segment (e.g., finding relevant content, joining communities).

2. Map current onboarding and pain points

Walk through the existing flow (sign-up, interest selection, feed, community discovery) and highlight where users drop off or get confused. Use data or reasonable assumptions to pinpoint friction.

3. Propose ML-driven improvements

Suggest 2-3 concrete enhancements that leverage ML (e.g., personalized interest selection, dynamic feed curation, community recommendations). Explain how each addresses a pain point and improves user experience.

4. Define success metrics and experimentation plan

Choose metrics (e.g., activation rate, time-to-first-engagement, D7 retention) and outline an A/B testing strategy with guardrail metrics. Mention offline evaluation and potential pitfalls.

Key Points to Mention

User segmentation: lurkers vs. contributors, niche vs. broad interests, and their distinct onboarding needs.
Current flow breakdowns: overwhelming interest selection, irrelevant default feed, difficulty finding communities.
ML techniques: collaborative filtering, content embeddings, contextual bandits for exploration/exploitation.
Prioritization criteria: impact on key metrics, implementation cost, and alignment with Reddit's values.
Success metrics: activation rate, time-to-first-upvote/comment, D1/D7 retention, and guardrails like content diversity.
Experimentation: A/B testing with proper randomization, sample size, and offline-online validation.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.