We are seeking a hands-on Agentic AI Architect with strong expertise in Agentic Workflows, MCP, C#, SQL Server, Angular, and Azure Cloud technologies to architect, design, build, and optimize enterprise-grade AI platforms. The ideal candidate will have hands-on experience with modern AI frameworks, including Agentic Workflows, MCP, Semantic Kernel, Kernel Memory, Azure AI Foundry, and the integration of multiple LLMs in advanced architecture such as Retrieval Augmented Generation (RAG).
This role requires a balance of advanced AI architecture experience, hands-on full-stack development skills with a forward-looking mindset on AI-driven platform development, ensuring scalability, security, and performance in production systems.
We are offering an opportunity to work with cutting-edge AI and cloud technologies.
This position offers flexibility for hybrid work schedules to include both in-office presence and telecommute/virtual work, to be based in Houston, TX or Dallas, TX.
Key Responsibilities
Platform Architecture & Technical Leadership
- Define and evolve the enterprise AI platform architecture, breaking down business and functional requirements into scalable, secure, and maintainable technical designs.
- Establish the long-term architecture runway for AI products, owning POCs, design spikes, and forward-looking evaluations of emerging technologies—including agentic architectures, orchestration frameworks, and tool-use patterns.
- Apply expertise in distributed systems and API design to ensure the platform is robust, observable, and simple to extend.
- Ensure best practices in secure coding, performance optimization, resilience, and lifecycle maintainability.
AI, Agentic Workflows & Cloud Integration
- Architect and implement enterprise-grade RAG (Retrieval Augmented Generation) pipelines, including vector search, embeddings, knowledge stores, and LLM orchestration.
- Design and operationalize agentic workflows, including multi-agent collaboration, tool-calling agents, planner/executor patterns, and automated reasoning loops.
- Build and integrate production-grade AI agents capable of interacting with external tools, APIs, enterprise systems, and knowledge sources.
- Utilize the Model Context Protocol (MCP) to expose internal tools, datasets, and enterprise APIs to LLMs in a secure, governed manner.
- Integrate and operationalize multiple LLM providers (Azure OpenAI, OpenAI, Anthropic, open-source models) into production systems.
- Leverage Azure AI services—Semantic Kernel, Kernel Memory, AI Foundry, Cognitive Search—to enable intelligent, context-aware platform capabilities.
- Design, deploy, and optimize cloud-native AI applications in Azure with emphasis on cost efficiency, observability, resilience, and security.
Cross-Functional Collaboration & Mentorship
- Partners with enterprise architects, product managers, developers, and data engineers in an Agile/Scrum environment to ensure architecture aligns with product strategy.
- Provide hands-on mentorship on AI development patterns—including agent-building, RAG troubleshooting, prompt design, vector search patterns, and orchestration frameworks.
- Conduct architecture reviews, code reviews, and design sessions to drive engineering quality and consistency.
- Translate business needs into scalable platform capabilities that support long-term product and enterprise roadmaps.
Innovation, Governance & Continuous Improvement
- Stay current on rapidly evolving AI trends: multi-agent systems, model orchestration, tool-use protocols, evaluation frameworks, prompt engineering, latency optimization, and AI safety approaches.
- Recommend improvements to development, testing, deployment, and estimation processes to increase delivery velocity while maintaining quality and compliance.
- Establish platform-level governance patterns including architectural guardrails, reusable modules, MCP tool adapters, and standardized agent templates.
- Champion experimentation, rapid prototyping, and continuous adoption of modern AI engineering practices to increase the organization’s AI maturity.