R

Full Stack Agentic AI Engineer

Robert Half
2 hours ago
Full-time
On-site
Richardson, Texas, United States
Full Stack Agentic AI Engineer

Increase your chances of reaching the interview stage by reading the complete job description and applying promptly.

We are looking for a

Full Stack Agentic AI Engineer

to join our

Agentic AI team . In this role, you will design and build scalable systems that power

conversational AI agents

and orchestrate agentic workflows. You’ll work across the stack—front‑end, back‑end, and cloud—while leveraging AI coding assistants and modern agentic frameworks to deliver enterprise‑grade solutions. A key part of your responsibilities will include

developing and automating evals

to rigorously test Agentic AI applications for performance, reliability, and safety.

About the Team Our Agentic AI team is pioneering

agentic AI development . We build extensive

Agentic AI solutions

spanning across multiple B2B and B2C use cases. Our mission is to deliver intelligent, autonomous systems that improve engagement, reduce operational overhead, and ensure compliance and quality.

What You’ll Do (Responsibilities)

Build and orchestrate AI agents

end‑to‑end: design tasks, tools, memory, planning, and guardrails; implement agent workflows using modern agentic frameworks (e.g., LangChain, Agents SDKs). Clear communication

and

cross‑functional collaboration

with Product, QA, Support, Design and our Customers to deliver outcomes; communicate technical decisions and trade‑offs clearly to both technical and non‑technical stakeholders. Develop with AI coding assistants

(e.g., Copilot‑style tools) to accelerate delivery while ensuring output meets

enterprise design, coding, security, reliability, performance, and compliance standards.

Ensure you do not over-engineer solutions. Manage coding agents. Eval first development

(offline/online) focused on Agentic AI application quality: define and build metrics and KPIs that showcase agent performance and business impact. Contribute to existing eval framework or standardizing evals across AI products. Architect for scale

across the stack (MCP Servers, API design, state/queueing, caching, streaming, observability, multi‑tenant isolation, cost controls) Translate product requirements into engineering designs

(specs, sequence diagrams, API contracts), drive design and architecture board reviews. Integrate agents with existing services and new microservices

(front‑end + back‑end), delivering highly

scalable and resilient

solutions Support in release and incident management and monitoring. Champion engineering excellence : code reviews, testing, accessibility considerations, performance optimization, and secure coding practices.

What You’ll Bring (Required Qualifications)

Full‑stack development:

Expertise in building

scalable back‑end services

using TypeScript/JavaScript/Python with strong API design and data modeling and developing

modern front‑end applications

using React (hooks, context, state management), including UI performance optimizations and accessibility best practices. Eval / Domain / Behavior Driven Development

(EDD/DDD/BDD) with focus on end user experience across all channels: UI, Chat, Messaging, Voice, Email, Mobile, etc. Hands‑on experience in building agents using

Agentic AI frameworks and Agents SDKs

including tool integration and persistence. Web technologies:

Experience with real-time communication using

WebRTC

and

WebSockets , including managing streaming connections and ensuring reliable data flow. Expertise in

cloud platforms

(AWS/GCP/Azure) and experience with

containerization

(Docker), orchestration (Kubernetes), CI/CD practices and infrastructure‑as‑code (e.g., Pulumi/CDK/Terraform). Strong

analytical and debugging skills

(logs, traces, performance tuning across front-end and back-end) paired with great attention to detail. Highly

skilled at designing structured prompts and prompt engineering techniques, tool‑augmented flows, and safety guardrails for coding assistants and agents. Exceptional ability to communicate complex technical concepts, drive consensus among both technical and non-technical stakeholders. Expert use of coding assistants

to build

enterprise‑grade, scalable

software—balancing speed with quality and standards. Agentic products shipped to production:

prior

end‑to‑end ownership

from prototype to production (observability, SLOs, incident handling). xsgimln Experience building online and offline evals for Agentic systems.

Preferred technology: TypeScript, Python, React, React Flow, OpenAI Agents SDK, MCP Server/Tools, GitHub Copilot, Anthropic Claude, OpenAI Codex, Serverless, Evals, Pulumi, Playwright