Software Engineer - Level 3

Hybrid role: Tues/Wed/Thurs in office at TX-Richardson-1680 N. Glenville Dr. Standard hours M–F 8 am–5 pm. Assignment runs through 2/28/2027 with possible extension and potential conversion to full-time employment.

About The Role

We're looking for an AI Engineer to design and build production-grade AI agents using Google's Agent Development Kit (ADK). You'll work on multi-agent systems that integrate with enterprise data platforms and business tools, taking agents from prototype through to reliable production deployment.

Responsibilities

• Designing and building AI agents and multi-agent workflows using ADK
• Agent orchestration patterns: tool use, function calling, sub-agent delegation, and state/session management
• Integrating agents with enterprise data sources (Snowflake, Salesforce, or similar) and internal APIs
• Deploying and scaling agents on Google Cloud (Vertex AI Agent Engine or equivalent)
• Evaluating and tuning agent performance, latency, and cost across different model backends
• Building tooling and guardrails around agent behavior: observability, testing, and safety checks
• Collaborating with data engineering and platform teams to expose reliable, well-governed data to agents

Requirements

• 8+ years of software engineering experience, with recent focus on applied AI/agent systems
• Hands-on experience building agents with Google ADK (or comparable agent frameworks such as LangGraph, AutoGen, CrewAI, with strong ability to ramp on ADK)
• Strong Python development skills; TypeScript/JavaScript a plus
• Experience with LLM APIs (Gemini, Claude, OpenAI, or similar) and prompt/tool design
• Understanding of agent architecture patterns: planning, tool orchestration, memory/state, multi-agent coordination
• Experience with REST APIs and backend service integration
• Familiarity with Google Cloud Platform (Vertex AI, Cloud Run, IAM, Secrets Manager)
• Experience shipping AI features to production, not just prototypes
• Comfortable working independently in ambiguous, fast-evolving technical territory
• Clear communicator, able to explain agent behavior and architecture to both technical and non-technical stakeholders

Skills

• Experience with Model Context Protocol (MCP) or other agent-to-tool integration standards
• SQL proficiency, ideally with Snowflake (CTEs, Dynamic Tables, Views)
• Experience with Snowflake Cortex / Cortex Analyst
• Familiarity with Salesforce data models (Account IDs, object relationships)
• Experience with evaluation frameworks for LLM/agent output quality
• Background integrating agents into existing enterprise applications (React/Next.js, Node.js/NestJS)

What Success Looks Like

• Agents that are reliable, observable, and behave predictably in production
• Clean integration between agents and underlying enterprise data/systems
• Reduced manual effort for end users through well-scoped agent automation