Senior Software Engineer - AI Pod

About Toast and AI Pod

Toast develops technology solutions aimed at helping restaurants and local businesses thrive in the digital era by streamlining operations, boosting sales, enhancing customer engagement, and improving employee satisfaction. The AI Pod team is an internal platform unit focused on artificial intelligence systems. They design the infrastructure, governance, and developer tools that enable AI to become a scalable and dependable productivity multiplier throughout the product development lifecycle. Their offerings include the LLM Proxy, AI Key Management, observability pipelines, a Claude Plugin Marketplace, and autonomous agents integrated within software development workflows.

Key Responsibilities

• Design and implement features across AI platform components such as the LLM Proxy, AI key management systems, and observability pipelines.
• Create and enhance autonomous AI agents involved in software development lifecycle processes, including tools like the AI Review GitHub App and Developer Platform MCP integrations.
• Contribute to and build plugins for the internal plugin marketplace that encode organizational architectural and quality standards into AI assistant capabilities.
• Develop MCP (Model Context Protocol) services and templated solutions to accelerate teams' ability to craft AI-driven workflows.
• Produce high-quality, well-tested code while actively participating in design discussions and code reviews.
• Engage cross-functionally with engineers, product managers, and platform teams to gather developer requirements and translate them into robust platform solutions.
• Support the team's adoption and advancement of agentic development methodologies and tooling.
• Utilize state-of-the-art AI tools to optimize personal development workflows, increase efficiency, and pioneer novel approaches, fostering an innovative and productive team environment.

Qualifications and Experience

• Bachelor's degree combined with a minimum of 5 years experience designing and deploying scalable backend systems.
• Proven background working with large language model (LLM)-powered agents or AI-assisted developer tools.
• Understanding of AI coding assistants (such as Claude Code, Cursor, GitHub Copilot) with enthusiasm for extending them via custom plugins, skills, or hooks.
• Familiarity with Model Context Protocol (MCP) or comparable frameworks supporting agentic integration.
• Proficient in prompt engineering and knowledge of how LLMs interpret context and instructions.
• Strong capabilities in Java, Kotlin, or other object-oriented programming languages.
• Experience with distributed systems architecture, RESTful APIs, and cloud-native infrastructure.
• Customer-focused mindset dedicated to improving developer experiences with dependable and well-engineered platform solutions.

Preferred Additional Skills

• Knowledge of Agent-to-Agent (A2A) protocols and orchestration patterns involving multiple AI agents.
• Background in machine learning concepts, model evaluation processes, or ML infrastructure.
• Experience developing developer experience platforms, internal tools, or API gateways.
• Familiarity with observability tools tailored for AI/LLM systems, such as Langfuse or DataDog.

Work Environment and Culture

This role requires onsite presence in Dublin, Ireland, two days per week as part of a hybrid working model. Toast fosters a culture centered on continual learning, innovation, and collaboration, empowering teams globally and regionally. They emphasize diversity, equity, and inclusion in all facets of their operations and hiring processes.

Additional Information

Toast offers competitive compensation and benefits designed to attract and retain top talent while supporting a healthy and flexible lifestyle. They are committed to providing accessible and inclusive hiring practices, including accommodations for disabilities upon request. AI tools are used to assist recruiters and interviewers by handling documentation tasks to enhance candidate conversations, but final hiring decisions are human-made.