Principal Forward Deployed Engineer

Sony Pictures Entertainment•Culver City, CA

About The Position

The Enterprise Intelligence Platform (EIP) builds and operates the AI platform that powers Sony Pictures Entertainment and drives AI Development Life Cycle (AI-DLC) adoption across the studio. This role is a newly created Principal Forward Deployed Engineer position within EIP. The individual will play a pivotal role in building the team, shaping the work delivered, and establishing new ways of partnering with the business to bring AI into production. This person will be deeply embedded with teams across Sony Pictures and help pioneer the AI-DLC that will underpin the next generation of solutions. This is a full-stack senior engineer role that partners with other parts of IT and with SPE business teams to ship full applications end to end, using the AI Development Life Cycle (AI-DLC) to compress the build cycle. Each engagement runs from scoping through prototype, production hardening, and handoff to the incumbent application team. Reusable patterns from every engagement flow back into the EIP platform. This person will report directly to the SVP of Enterprise Intelligence Platform. This is a write, ship, and own role, focusing on production code, AI-DLC practice, and platform feedback, not a consulting or solution-architect role. Some applications will include LLM inference at runtime (RAG, agents, prompt-based interactions).

Requirements

Full-stack production engineering experience across the whole application stack: object-oriented development proficiency, backend services, data pipelines, user-facing application surfaces (web UIs, internal tools, integrations), cloud (AWS and / or Azure), containers, and infrastructure-as-code.
You ship every layer, not just the AI components.
Hands-on experience using AI-assisted development tooling (Claude Code, Cursor, GitHub Copilot, Windsurf, or equivalent) to ship production code at meaningfully faster cycle times than traditional development.
Comfort generating, reviewing, and iterating on AI-generated code at velocity, with the judgment to know what to keep and what to throw away.
8+ years of professional software engineering experience.
Some production experience with LLM-based systems (RAG, agents, prompt architecture, evaluation frameworks), applicable to the subset of engagements where the application includes runtime AI.
A track record of satisfying enterprise security, governance, and data-handling requirements when shipping software in regulated or rights-sensitive domains.
Strong stakeholder communication. You can sit with a non-technical business partner, draw out what they actually need beyond the surface ask, propose a solution with them, and own translating it into a decomposed, shippable plan in real time. The same skill carries from frontline operators through senior executives.
Comfort operating with high autonomy in ambiguous, fast-moving engagements. You write, ship, and own production code; you do not hand off recommendations.
Bachelor's degree or equivalent practical experience.

Nice To Haves

Prior Forward Deployed Engineer or embedded senior engineer experience.
Deep experience with AI-assisted development workflows in a production setting (not just experimentation).
Experience coaching teams through methodology shifts (DevOps adoption, agile transformations, AI-assisted development rollouts).
Prior experience in media, entertainment, or other rights-sensitive industries.
Experience with the EIP foundation platforms (Amazon Bedrock, Azure OpenAI, Anthropic, OpenAI APIs) or with Moveworks / Agentic Foundation.
Cloud platform certifications (AWS, Azure, or GCP) and / or production-AI / MLOps certifications.

Responsibilities

Embed with IT and business teams on priority SPE projects and ship full applications end to end (data layer, backend services, AI integration where the application needs it, user-facing surface), using AI-DLC to compress the development cycle dramatically.
Lead AI-DLC adoption in your partner teams: coach the team into team-tier self-sufficiency on AI-assisted development so they continue using the practice after you roll off.
When the application requires inference at runtime, build production LLM components on the EIP platform (RAG, agentic orchestration, prompt architecture, evaluation frameworks, observability, and guardrails).
Work directly with non-technical business stakeholders, from frontline operators and analysts through senior executives, to draw out what they actually need, shape a solution together, and translate it into a decomposed, shippable plan in real time.
Harvest reusable patterns from every engagement back into the EIP platform backlog so what you build becomes platform capability and successive engagements compound.
Hold the line with senior stakeholders when the request needs to change, and own what you ship through the handoff to the incumbent application team.