← reflectionai / Product Manager - OSS
tailored_resume_v2 / art_K1tLI8bMDJk
↓ Download .docx ↓ Download .pdf PDF requires LibreOffice installed
What changed for reflectionai
| change | why it matters |
|---|---|
| Summary rewritten to lead with developer platform ownership and OSS/AI model evaluation credentials | JD's first requirement is shipping developer-facing products in OSS or AI and owning the full product surface — the 675M+ engagements and RL workbench are the strongest proof points |
| Intuit reordered to lead Experience section (above Streamio) | Intuit is the highest-relevance role (score 5) with the most direct proof of developer platform ownership, ecosystem metrics, and public forum credibility that Reflection is screening for |
| Intuit bullets reframed to emphasize downstream adoption, ecosystem dynamics, success metrics instrumentation, and CTO-level research/strategy shaping | JD explicitly screens for understanding OSS success beyond stars/downloads — contribution health, downstream adoption, ecosystem dynamics — and for changing direction based on engineer feedback |
| Added explicit 'changed product direction based on engineer/researcher input' framing to Intuit Service Language Assessment bullet | JD states 'You have changed your mind because of feedback from a researcher or engineer and can describe the specifics' — this is a screening criterion |
| RL Workbench moved to lead the Projects section | Directly maps to Reflection's core work on open-weight model post-training; benchmarking TRL/VeRL/OpenRLHF/NeMo RL is the strongest research-PM credibility signal |
| RL Workbench second bullet reframed to include 'directly informing what to cut and when in post-training pipelines' | JD explicitly asks for PMs who can talk about 'what gets cut and when' — this maps the benchmarking work to that decision-making lens |
| aeval bullets reframed to emphasize 'defining and instrumenting success metrics that go beyond surface-level outputs' | JD requires 'Define and instrument success metrics that tell you whether the bet is working, not just whether you shipped' — aeval is direct proof |
| Streamio MCP SDK and OpenClaw bullets reframed to emphasize developer platform surface and OSS tooling exposure | JD requires owning developer platforms and ecosystem relationships; MCP SDK integration is the strongest OSS-adjacent proof point from Streamio |
| Fintellect condensed to 2 bullets | Lower relevance to OSS/model release focus; space allocated to higher-signal roles and projects |
| Kaiser and IBM condensed to minimum bullets | Low relevance to OSS PM role; retained for completeness and career arc but not padded |
| DeveloperWeek 2022 and Splunk .conf speaker credentials surfaced prominently in Intuit and Splunk bullets | JD requires ability to 'hold your own externally with senior partners and in public forums' — conference speaking is direct evidence |
JD analysis (20 key phrases)
Key phrases: open-source release end-to-endcontribution healthdownstream adoptionecosystem dynamicscommercial reciprocitymodel releasesdeveloper-facing productspartner and ecosystem relationshipsGTM motiondeveloper adoptionopen weight modelsresearch leadsproduct directionlicensing, documentation, communityinference providerscloud partnersdeveloper platformssuccess metricswhat gets cut and whenopen superintelligence
Hard requirements:
- Shipped model releases or developer-facing products in OSS or AI
- Can speak specifically to what went wrong, what was cut, and what metrics actually did
- Understands OSS success beyond stars/downloads: contribution health, downstream adoption, ecosystem dynamics, commercial reciprocity
- Can hold own externally with senior partners and in public forums
- Has changed mind based on researcher/engineer feedback
- Owns product end-to-end: research shaping, surface, OSS release, community, ecosystem, GTM
Preferred qualifications:
- Background in AI/ML research or research-adjacent product work
- Experience with open-weight model releases
- Developer platform and ecosystem partnership experience
- Go-to-market ownership for developer/AI products
- Ability to define and instrument success metrics
Per-role mapping (9 roles scored)
| role | score | reframe angle | JD phrases that map |
|---|---|---|---|
| Streamio AI — Founder & CEO | 4/5 | End-to-end AI product ownership with developer-facing SDK/tooling and 0-to-1 GTM execution | developer-facing products, GTM motion, developer adoption, what gets cut and when, open-source release end-to-end |
| Fintellect AI — Founder & CEO | 3/5 | AI product GTM and multi-model orchestration ownership | GTM motion, developer adoption, model releases |
| Intuit — Staff PM, Developer Frameworks & Platform Infrastructure | 5/5 | Developer platform ownership at enterprise scale — SDK tooling, DevPortal, ecosystem metrics, public speaking | developer-facing products, developer platforms, downstream adoption, ecosystem dynamics, success metrics, developer adoption, cloud partners, licensing, documentation, community |
| Splunk — Senior PM, Search Orchestration | 3/5 | End-to-end microservice product ownership with external community presence | developer-facing products, what gets cut and when, success metrics, partner and ecosystem relationships |
| Kaiser Permanente — SOA Technical PM | 2/5 | Platform infrastructure ownership at scale | downstream adoption, success metrics |
| IBM — Software Engineer | 1/5 | Engineering foundation supporting research-PM credibility | — |
| RL Workbench | 5/5 | Research-grade RL post-training platform directly relevant to open-weight model development | model releases, open weight models, research leads, downstream adoption |
| aeval — AI Model Evaluation Platform | 5/5 | Model evaluation infrastructure — directly maps to defining and instrumenting success metrics for model releases | model releases, success metrics, contribution health, downstream adoption |
| BRAIN — NeurIPS 2014 | 4/5 | Published AI researcher with hands-on model development — earns credibility working directly with research leads | research leads, model releases, open weight models |
Tailored summary
Technical PM with 12+ years owning developer-facing platforms end-to-end — from SDK tooling and DevPortal ecosystems to AI model evaluation and RL post-training infrastructure. Scaled Intuit's developer platform to 675M+ engagements and 50K TPS; built and benchmarked GRPO/DPO post-training pipelines across TRL, VeRL, OpenRLHF, and NeMo RL. NeurIPS-published AI researcher who has shipped 0-to-1 AI products, led GTM motions without hand-offs, and spoken publicly at DeveloperWeek and Splunk .conf on developer platform strategy.