Product Development Engineer I (AI-Native) (Onsite role)

Phenom People Pvt Ltd•Lower Gwynedd Township, PA

35d•$80,000 - $90,000•Onsite

About The Position

Our purpose is to help a billion people find the right work! Phenom is an AI-Powered talent experience platform that is redefining the HR tech space. We have grown into a global organization with offices in 6 countries and over 1,500 employees. As an HR tech unicorn organization, innovation and creativity is within our DNA. Come help us make every talent moment Phenomenal! We are looking for a Product Development Engineer (AI-Native) who takes ideas and customer needs and turns them into working, validated product features and agentic capabilities — owning them from first prototype to production. Here, everyone wears multiple hats: the same person who builds a feature also ships it, supports the people using it, and debugs issues directly. You own outcomes end-to-end, not a narrow slice.

Requirements

1-4 years of software development experience.
Bachelor's degree in Computer Science, Software Engineering, Information Technology, or a related technical field required; Master's degree preferred.
Strong fundamentals and critical thinking — AI raises the bar on judgment, it doesn’t replace it.
Comfortable across the stack and building with modern AI coding tools and agents.
Customer empathy — you can sit with a customer, ask the right questions, and decide which problems are worth solving.
End-to-end ownership — you finish what you start and keep improving it.
Adaptable — eager to adopt new tools and reinvent your workflow as the tooling changes (often weekly).
A strong communicator who aligns quickly with colleagues and customers.
You thrive on variety — building, debugging, shipping, and customer calls in the same week.
Programming proficiency — write clean, maintainable code in at least one modern language (e.g., Python, Java, JavaScript / TypeScript, or Go).
Computer science foundations — strong understanding of data structures, algorithms, complexity (Big-O), and decomposing problems into clean, testable components.
Software design principles — object-oriented and functional concepts, clean-code practices, and sensible code and API design.
Version control & collaboration — day-to-day fluency with Git, branches, pull requests, and code review.
APIs & databases — working knowledge of REST APIs and JSON, plus basic SQL and data modeling with relational and/or NoSQL stores.
Testing & debugging — writing unit and integration tests and diagnosing issues methodically from logs, traces, and stack traces.
Command line & cloud (a plus) — comfort in a Unix / Linux shell, with exposure to a cloud platform (AWS, Azure, or GCP) and CI/CD concepts.
AI tooling (a plus) — hands-on with modern AI coding assistants and familiarity with LLM or agent concepts.

Nice To Haves

Bonus: experience building AI / agent products, workflow orchestration, observability, or building agent workflows on an Agent SDK.

Responsibilities

Build innovative products on the Phenom platform — prototype fast, then harden what proves valuable into something customers can rely on.
Take ideas to production — design, build, and validate features end-to-end, then keep improving them after they ship.
Build and improve reusable agentic skills — packaged, versioned capabilities anyone in engineering can compose and re-run, so the next build starts from a proven asset, not a blank page.
Practice AI-assisted, increasingly agentic engineering — drive PRs with coding agents and your own custom skills, and wire each PR to also update the observability it depends on.
Run a controlled debugging loop — work from logs, traces, and prior incidents to find and fix root causes quickly.
Own delivery to customers — deployment gates and hypercare, partnering with reliability engineers on deep platform issues.
Keep rich context for agents — maintain the structured, per-customer context that every agent and skill draws on before it acts.
Respect the guardrails — analysis is separated from action, high-risk changes require human approval, and every decision is logged for audit.