← anthropic / Product Manager, Developer Productivity
candidate_questions / art_bORkt0pG_MA
role
model
anthropic/claude-sonnet-4.6
created
2026-05-19T20:51
Interviewer
The interviewer is a PM at Anthropic (listed as 'Test PM at AnthroDemo' — likely a placeholder or internal alias), with prior experience at Stripe and a Stanford CS background who has built developer platforms. Given the Stripe background, they likely have deep familiarity with developer-facing products, API design, and platform thinking at scale. As a peer PM, the conversation will likely focus on product craft, how Felix thinks about developer experience, prioritization trade-offs, and his thesis on AI-native development. Shared context includes developer platform building — Felix's Intuit work on SDK Starter Kits, ICE, and DevPortal maps closely to the interviewer's Stripe developer platform experience.
Questions to ask them (20)
| category | question | why |
|---|---|---|
| interviewer_experience | You came from Stripe, which is legendary for its developer experience bar — the 'it just works' philosophy that shaped how a generation of engineers think about APIs. What did you bring from that mindset when you joined Anthropic, and where did you have to unlearn or adapt it for an internal platform context? | Anchors on their Stripe background, signals Felix knows Stripe's developer culture deeply, and surfaces how internal vs. external platform thinking differs at Anthropic — directly relevant to the role. |
| interviewer_experience | Having a Stanford CS foundation and then moving into PM — especially on developer platforms — how much do you find yourself going deep into the technical architecture of build systems or CI infrastructure day-to-day, versus staying at the product abstraction layer? I'm curious how that balance has evolved for you at Anthropic. | Reveals how technical the PM role actually is in practice, and lets the interviewer share their working style — useful for Felix to calibrate how to position his own technical depth. |
| interviewer_experience | What's kept you at Anthropic? I imagine someone with your background gets recruited constantly — what's the thing that makes you stay and keep building here? | Genuine retention signal and culture probe — the answer reveals what's actually energizing about the environment beyond the mission statement. |
| role_team_dynamics | If I'm in this role and it's 90 days in — what does 'good' look like? Not just ramped-up, but actually having made a meaningful contribution that the team notices? | Classic 30/60/90 probe but pushed to 90 days for specificity — reveals what the team actually needs most urgently and what success signals matter to this interviewer. |
| role_team_dynamics | The JD mentions partnering across Infrastructure, Inference, Research, and Product Engineering — that's a wide stakeholder surface. In practice, which of those relationships tends to be the most friction-generating, and how does the current team navigate it? | Surfaces real organizational dynamics and where the PM will need to spend political capital — avoids being blindsided post-hire. |
| role_team_dynamics | How is the Developer Productivity PM function structured today — is this a standalone PM role embedded in an eng team, or is there a broader product org with multiple PMs working adjacent problems? And who does this role report into? | Clarifies org structure, reporting lines, and whether Felix would be a solo PM or part of a team — critical for evaluating scope and support. |
| role_team_dynamics | What's the single biggest developer productivity pain point that the team hasn't been able to crack yet — the thing that keeps coming up in eng retros or leadership reviews? | Reveals the hardest unsolved problem, which is where Felix would likely spend his first year — and tests whether the interviewer trusts him enough to be candid. |
| technical_environment | The JD mentions build and CI infrastructure across multiple cloud providers. At Intuit I worked across GCP and AWS in a monorepo context — can you give me a sense of the current build system landscape at Anthropic? Are you on Bazel, something custom, or a mix? And where are the biggest reliability or speed pain points today? | Directly leverages Felix's Intuit monorepo and multi-cloud experience to ask a credible technical question — signals he can hit the ground running and isn't asking from scratch. |
| technical_environment | The accelerator toolchain piece — GPU, TPU, Trainium — is something I've been thinking about a lot, especially the developer experience challenges when researchers need fast, reproducible builds on compute that has strict compatibility constraints. How mature is the current toolchain abstraction layer, and is the PM expected to drive the vision there or more to coordinate between eng leads who own it? | Shows Felix has read the JD carefully and has genuine context on ML compute DX — probes the actual PM scope on a technically complex area. |
| technical_environment | How does Anthropic currently measure developer productivity? Are you using DORA metrics, SPACE, something custom — and how has the team started thinking about what those metrics mean when Claude is writing a meaningful share of the code? | Directly relevant to the JD's emphasis on AI-native productivity metrics — and signals Felix has a point of view on this evolution, not just awareness of the frameworks. |
| culture_working_style | When a PM and an engineering lead genuinely disagree on prioritization — say, velocity versus reliability — how does that get resolved here? Is there a clear decision-making framework, or does it tend to escalate, or does it depend on the relationship? | Reveals how much autonomy the PM actually has versus how consensus-driven or eng-led the culture is — critical for someone coming in to own a roadmap. |
| culture_working_style | Anthropic describes itself as a highly collaborative, single-cohesive-team research org. For a developer productivity PM whose customers are internal engineers and researchers — how do you balance being a service org that responds to requests versus having a strong product vision that sometimes says 'no, we're not building that'? | Probes the internal platform PM tension between being a service org and having strategic conviction — and whether the culture supports the latter. |
| culture_working_style | How much of this role is greenfield strategy versus inheriting an existing roadmap and system? I want to understand how much of the job is 'define what developer productivity means here' versus 'execute against a direction that's already been set.' | Directly evaluates the actual autonomy and ambiguity level — Felix has 0-to-1 founding experience and wants to know if that's valued or if this is more execution-mode. |
| growth_development | For a PM coming into this role with a strong technical background — ML research, build systems, platform infra — is there room to go deeper on the AI safety and interpretability side of Anthropic's work, or does developer productivity stay pretty siloed from the core research mission? | Explores whether Felix can grow toward Anthropic's core research mission over time, not just stay in an internal tooling lane — important for long-term fit. |
| growth_development | You went from Stripe to Anthropic — both high-bar engineering cultures but very different contexts. What's been the steepest part of your own learning curve here, and what do you wish someone had told you before you started? | Invites the interviewer to share their own growth experience, builds rapport, and surfaces real onboarding challenges Felix should anticipate. |
| strategy_vision | The JD talks about a fundamental shift — from AI as autocomplete to AI as autonomous collaborator. What's Anthropic's current internal posture on that shift for its own engineering org? Is Claude already doing meaningful autonomous work in the inner loop, or is that more of a 12–24 month horizon you're building toward? | Tests whether the vision in the JD is aspirational or already underway — and reveals how much of the role is building for a future state versus managing a present reality. |
| strategy_vision | One of the hardest problems in AI-native developer tooling is governance — how do you let teams safely delegate work to autonomous agents without creating security, compliance, or quality debt? Has Anthropic started building a framework for that internally, or is defining that governance model a core part of what this PM would own? | Shows Felix has thought deeply about the governance challenge specifically called out in the JD — and probes whether there's existing structure or a blank slate. |
| strategy_vision | On a 2–3 year horizon, what does 'Anthropic is the most productive place in the world to build frontier AI' actually look like in practice? What's the north star metric or experience that would tell you the team got there? | Invites the interviewer to articulate the vision concretely — and lets Felix assess whether the ambition matches his own thinking and energy. |
| shared_context | At Intuit I ran a Service Language Assessment across 9 languages to inform strategic investment decisions — essentially a platform rationalization exercise across a large eng org. At Stripe you were presumably thinking about similar questions around developer experience consistency at scale. I'm curious: when you're rationalizing a fragmented internal tooling landscape, how do you build the conviction to deprecate something that engineers love but that doesn't fit the long-term platform direction? | Draws on Felix's Intuit experience and the interviewer's Stripe background to have a peer-level conversation about a real platform PM challenge — builds credibility and rapport simultaneously. |
| shared_context | I built the ICE Self-Service platform at Intuit — reducing developer onboarding from weeks to minutes — and one of the hardest parts wasn't the build, it was adoption. Engineers had muscle memory around the old way. You've built developer platforms too — what's your playbook for driving internal adoption through product quality rather than mandate, especially when the old tool is 'good enough'? | Directly references a shared challenge (internal platform adoption) that both Felix and the interviewer have faced — positions Felix as a peer with hard-won experience, not a candidate asking for advice. |
Conversation starters
- I saw you came from Stripe — I've always admired how Stripe set the bar for what developer experience documentation and onboarding could feel like. I'm curious whether that 'docs as product' philosophy translated into how you think about internal developer platforms here.
- Your Stanford CS background plus Stripe plus Anthropic is a pretty specific arc — it feels like someone who's always been drawn to the infrastructure layer that other builders depend on. That resonates with me — a lot of my career has been about making the platform invisible so the engineers on top of it can move faster.
- I was thinking about the governance problem for AI agents in the inner loop before this conversation — specifically how you draw the line between 'Claude can do this autonomously' and 'a human needs to be in the loop here.' I'd love to hear how you're thinking about that at Anthropic right now.
⚠ Handle carefully
- The interviewer's profile is sparse ('Test PM at AnthroDemo') — this may be a placeholder or anonymized profile. Avoid referencing specific tenure, team names, or project details that aren't confirmed. Stick to what's stated: Stripe background, Stanford CS, developer platforms.
- Don't probe on why they left Stripe or imply Anthropic is a step down or lateral — Stripe is a prestigious company and the move to Anthropic is clearly a deliberate choice toward mission-driven work; treat it as an upgrade, not a departure.
- Avoid implying that Anthropic's internal tooling is immature or behind — the JD signals ambition, not dysfunction. Frame questions around 'where you're headed' rather than 'what's broken today' to avoid sounding critical of what the team has built.