Lead AI Engineer

Toshiba Global Commerce Solutions is seeking an AI Architect who is an expert in designing and delivering complex systems, acting as a go-to technical authority in the Agentic AI organization. The AI Architect will typically lead multiple projects or an entire critical domain of the platform. For example, they might own the “AI Code Execution and Validation” domain end-to-end, which includes everything from sandboxing, static analysis, runtime monitoring, to deployment integration for AI-generated code.

In this role, an engineer has broad autonomy to define technical approaches and is trusted to make decisions that affect the architecture of the overall system. They coordinate closely with the engineering manager and principal engineers to ensure the platform meets strategic goals. This level is analogous to advanced staff engineers at leading companies – those who have designed large-scale AI or distributed systems and are recognized for solving hard technical problems, such as multi-agent orchestration and cloud scalability.

 

Key Outputs & Outcomes:

• Strategic Technical Plans: Clearly defined architecture roadmaps and design documents for major parts of the platform, ensuring the team is moving in a coherent direction that aligns with business needs and technological evolution. Outcomes include approved architecture proposals and funded projects based on your plans.
• Successful Delivery of Complex Projects: Completion of large-scale projects under your technical leadership, on time and within performance/scope goals. For example, a new AI-driven code verification service is delivered and reduces human code review effort by, say, 50% – a direct measurable improvement tied to your initiative.
• Technical Innovation: Introduction of new technologies or approaches that significantly improve the platform’s capabilities or efficiency. This could be measured by performance metrics (e.g., system throughput improved by a factor due to your prototyped optimization) or capabilities (e.g., the platform can now support an additional programming language or framework for code generation, expanding Toshiba’s product offerings). Possibly also patents or internal IP generated from innovative solutions.
• Organizational Learning & Standards: The broader org benefits from best practices you establish. For instance, coding standards or security practices you champion become standard operating procedure, reducing bugs and incidents. Other engineers step up with improved design thinking and autonomy, influenced by your mentorship (e.g., you see mid-level engineers confidently leading sub-projects using approaches you advocated).

Responsibilities:

• Technical Roadmapping & Design:
  • Work with Principal Engineer and leadership to define the technical roadmap for your domain (e.g., “AI Code Execution and Validation”). Break down the long-term vision into concrete projects and architecture updates. For instance, plan a series of enhancements over the next year to enable the AI agent to handle larger codebases or multi-module systems autonomously. Author high-level architecture proposals and lead technical planning meetings.
• Project Technical Leadership:
  • Lead the execution of major engineering initiatives. This includes driving project timelines, making critical design decisions, and ensuring cohesion across components. For example, if rolling out a new autonomous code analysis service, decide on the tech stack (perhaps adopting an open-source static code analysis tool and extending it), outline the API contracts between services, and ensure that engineers on the project implement according to specifications. Perform risk management by identifying potential technical pitfalls early (scalability, security) and mitigating them.
• Innovation & Prototyping:
  • Investigate emerging technologies or methods to keep improving the agentic AI platform. Prototype solutions to particularly challenging problems. E.g., evaluate a new LLM with better code understanding, or prototype a new approach for the AI to self-debug runtime errors using reinforcement learning. Based on results, make recommendations (or decisions) on whether to integrate these innovations. This may involve reading research papers or collaborating with research teams, and turning insights into proof-of-concept code..
• Cross-Functional Collaboration:
  • Interface with other groups (Product, QA, Security, Operations) at a senior level. For instance, work with Product Management to understand requirements for new agent capabilities (like supporting a new