The mandate: End-to-end ownership of core product areas — voice agent behavior, onboarding, testing, and AI-driven automation. Strategy, roadmap, and success metrics grounded in real customer pain, adoption, and business impact. Scope spans real-time audio, telephony workflows, agent behavior, and developer tools. The JD calls it "highly technical with a wide scope," and that's accurate — this is not a feature-area PM seat. The strongest signal in this posting is the customer research expectation, and it's unusually concrete. The JD names specific behaviors: spending time with users, reviewing call transcripts, joining onboarding and support sessions, engaging customers directly to find unmet needs and failure points. Transcript review in particular is a real tell — Retell wants a PM who reads what actually happened on the calls rather than reasoning from dashboards. Recruiters should screen for this directly; it's the fastest way to separate real customer-obsessed PMs from ones who say it. What You'll Own
-
Customer outcomes for core product areas — voice agent behavior, onboarding, testing, AI-driven automation. Strategy, roadmap, and success metrics.
-
Deep customer understanding — user time, call transcript review, onboarding and support sessions, direct engagement
-
Outcome-driven PRDs — customer problem, requirements, end-to-end journey, solution concepts. Used to align engineering, design, and stakeholders on scope, tradeoffs, and success criteria before execution.
-
Competitive and market insight feeding strategy and prioritization
-
Scalable solution design with engineering, weighing customer value against time-to-impact and platform health
-
Cross-functional input from Sales, CX, and Customer Success — buying signals, objections, rollout blockers, expansion drivers
-
Design partnership from early problem framing and low-fidelity sketches
-
Execution: planning, sequencing, tracking, clear ownership
-
Early validation via prototypes, betas, and rapid iteration
-
Product rituals — roadmap planning grounded in customer evidence, prioritization frameworks tied to impact, clear communication of tradeoffs Requirements Hard gates:
-
5+ years product management, ideally B2B and product-led growth
-
Customer-obsessed in practice — regularly talks to users and reviews real usage: calls, sessions, workflows
-
Strong cross-functional collaboration with engineering, design, and sales
-
Founder-level ownership and bias toward action
-
Highly organized and execution-oriented — creates clarity in messy spaces
-
Technically curious and willing to go deep
-
Redwood City (confirm on-site status) Profile:
-
4–10 years, senior enough to own strategy without a director above them
-
Energized by ambiguity and fast iteration
-
Writes well — PRDs are named as the primary alignment artifact Strong bonuses:
-
Voice AI experience — "a strong advantage" in Retell's own words. Rarest and most valuable.
-
AI product experience generally — evaluation, non-determinism, agent behavior
-
PLG product experience — named explicitly; Retell is developer-first with self-serve motion
-
Developer tools or API product management — developer tools are in scope
-
Telephony, real-time systems, or CCaaS domain knowledge
-
Contact center or CX operational understanding
-
Has established product process at a company that didn't have one Location and visa:
-
Redwood City
-
Sponsorship: H-1B, TN, L-1, E-3, F-1 (OPT/CPT), and O-1 Anti-patterns
-
PMs who work through a research or CS layer and never touch customers. The JD's transcript-review language rules them out explicitly.
-
Roadmap and backlog managers with no strategy ownership
-
Big-company PMs with narrow feature-area scope and a support apparatus
-
PMs who can't go technical — real-time audio and telephony are in scope
-
Anyone needing a defined process to operate in; they're building the rituals
-
Pure enterprise PMs with no PLG or self-serve experience
-
Program managers with PM titles
-
Remote requirements Who Will Thrive Here
Someone who reads the transcripts. They've owned a technical B2B product end to end, they write PRDs people actually use to make decisions, and they'd rather find the failure point themselves than have it summarized. They can hold a roadmap conversation with the CPO and a latency conversation with an engineer in the same hour. At 50 people with a wide scope and real platform depth, they get to decide what a category-defining product looks like.