AI Architect

Panasonic Automotive•Peachtree City, GA

8d•Hybrid

About The Position

Panasonic Automotive is seeking a hands-on AI Solution Architect who combines architectural expertise with AI engineering to design and deliver innovative, production-grade enterprise solutions. The role involves analyzing complex business problems, identifying where AI is appropriate versus conventional technologies, and translating architectural decisions into implementations. This is an individual-contributor position that spans systems thinking, solution design, architecture decisions, experimentation, coding, reference implementations, and production delivery. The architect will partner with business and technical stakeholders to develop end-to-end enterprise AI solutions, integrate them with enterprise systems and data, and support adoption and operational handoff. The role also contributes reference architectures and templates and supports partner teams through design and code reviews.

Requirements

7+ years of software engineering experience, including 2+ years shipping LLM, RAG, or agent-based systems to production.
Strong Python skills and working knowledge of TypeScript or another backend language.
Hands-on experience with LLM APIs and open-weight models, function calling, structured outputs, and agent frameworks such as LangGraph or Semantic Kernel.
Experience with vector or hybrid search and embeddings.
Experience with MCP and tool-integration design; familiarity with A2A or similar agent interoperability standards.
Experience building LLM evaluations and observability, such as OpenTelemetry-based tracing, Langfuse, or LangSmith.
Cloud deployment experience on AWS, Azure, or GCP, with Docker, Kubernetes, and secure API design using OAuth/OIDC.
Daily use of AI coding assistants and agents in personal engineering workflow.
Experience delivering enterprise solutions that integrate with enterprise applications, SSO/identity, and data platforms under security and architecture review.
Clear communication of design trade-offs, both in writing and in person.
Valid authorization to work in the U.S.
Ability to perform the essential functions of the job with or without reasonable accommodations.

Nice To Haves

Model serving or fine-tuning experience, such as vLLM or LoRA, including edge or on-device inference.
Working knowledge of AI governance frameworks such as NIST AI RMF, ISO/IEC 42001, or the EU AI Act.
Automotive, embedded, or manufacturing software background.
Vendor-neutral certifications or public technical work in AI systems.

Responsibilities

Decompose ambiguous business problems into capabilities, information, behaviors, system boundaries, interactions, dependencies, constraints, and quality attributes, and translate them into coherent end-to-end architectures.
Apply architecture principles, patterns, quality attributes, and structured decision-making; evaluate alternatives and communicate decisions, alternatives, and trade-offs.
Challenge assumptions and develop pragmatic architectures, deciding when to use AI, deterministic technologies, or a combination of both.
Work with business and technical stakeholders to turn use cases into enterprise AI solutions, including requirements, solution design, build-versus-buy and platform choices, and delivery roadmaps.
Integrate AI with enterprise systems and data, including ERP, identity and SSO, data platforms, APIs, and legacy applications, in line with architecture, security, and compliance standards.
Design and build single- and multi-agent systems, including orchestration, tool use, memory, and human-in-the-loop controls.
Build RAG and context-engineering pipelines, addressing retrieval quality, long-running agent memory, and token and latency management.
Build MCP servers and agent integrations that expose enterprise tools and data through governed, least-privilege access.
Run LLM access through a gateway with model routing, fallback, quotas, and cost and usage tracking.
Build evaluation, tracing, and monitoring for LLM and agent behavior, including regression tests and release gates.
Secure agents against prompt injection, data leakage, and over-permissioned tools, and maintain audit trails of agent actions.
Deploy AI services using containers, CI/CD, and infrastructure as code, and troubleshoot them in production.
Support enterprise rollout of AI solutions through pilots, user adoption, operational handoff, and feedback loops.
Publish reference architectures and working templates, and review designs and code with partner teams.