Applied AI Engineer

About the Role

Own Silvia's AI platform end-to-end - from agent architecture to production deployment to monitoring and iteration. Design and orchestrate multi-agent systems where specialized agents collaborate - portfolio analysis, market monitoring, spending patterns, risk assessment - each with its own tools, context, and guardrails.

Our values are simple and the lifeblood of our company:

• How you do one thing is how you do everything
• The team you build is the company you build.
• Compete, don't complain.
• Think in decades, act in days.
• Unleash your child-like curiosity.

Our culture is unique, but if you want to build the future alongside other A+ colleagues in a highly entrepreneurial environment, there's no better place to be.

What You'll Do

• Implement Model Context Protocol (MCP) to connect agents to live data sources - databases, Plaid, market APIs - in a structured, composable, production-grade way
• Build context-enriched AI systems that ground model outputs in real user data (portfolios, transactions, market feeds) so responses are accurate and personalized, not generic
• Design and ship agentic workflows that monitor portfolios in real-time and surface timely recommendations - rate hike impacts, sector rotation signals, anomalous transactions, spending patterns
• Build the AI chat experience - enabling users to have natural, grounded conversations with their financial data
• Own AI infrastructure - model serving, inference optimization, cost management, and scaling to support rapid user growth while maintaining fast response times
• Develop evaluation, observability, and guardrail frameworks - define quality benchmarks, track hallucination rates, measure agent reliability, and continuously improve accuracy
• Build and maintain the data pipelines that feed AI systems - ingesting and processing real-time financial data from thousands of account integrations
• Collaborate closely with product and engineering to turn model capabilities into features users love

What You Bring

• 7+ years of software engineering experience, with at least 2+ years building LLM-powered products in production environments - not prototypes, not demos, production
• Proven track record designing and shipping multi-agent systems - you've built specialized agents that collaborate, handle failures gracefully, and run reliably at scale
• Experience with Model Context Protocol (MCP) or similar standards for connecting AI models to live tools, APIs, and data sources in a composable, production-grade way
• Deep expertise in prompt engineering, tool-use design, and agent orchestration - you've made agents that are trustworthy enough for high-stakes decisions, not just impressive in a demo
• Experience building context-enriched AI systems over user data - you know how to ground model outputs in real, structured data so responses are accurate and personalized, not generic
• Strong engineering fundamentals - you write clean, tested, well-architected code. AI doesn't excuse bad software engineering
• Built evaluation, observability, and guardrail frameworks for agent systems - you can quantify when agents are wrong and know how to fix it
• Comfortable owning AI infrastructure - model serving, cost optimization, and scaling inference in production
• Interest in finance, investing, or personal finance - domain expertise is a plus but not required