Agentic AI Developer

Overview: We are seeking a Senior Agentic Application Developer to architect, build, and operate production-scale, consumer-facing agent systems. This role is for engineers who have moved beyond prompt experimentation and are comfortable designing stateful, observable, fault-tolerant agent architectures that run reliably in real products. You will own agent design end-to-end—from orchestration and schema design to evaluation, deployment, and long-term operability.---Core Responsibilities· Architect stateful agent systems using Lang Graph, including complex graph topologies, branching logic, retries, and long-lived state· Design strict, versioned agent schemas using Pydantic to enforce contracts between agents, tools, and downstream systems· Build and operate tool-using agents that safely orchestrate calls across APIs, databases, and internal services· Implement robust failure handling for agents, including timeouts, partial execution recovery, fallback strategies, and human-in-the-loop escalation· Instrument agent execution using Lang Smith, defining custom traces, evaluations, and performance benchmarks· Optimize agent systems for latency, cost, and reliability at production scale· Lead the design of agent memory strategies (short-term, long-term, episodic, and externalized memory)· Own agent deployment pipelines, including versioning, rollbacks, and safe rollout strategies· Partner with product and platform teams to translate complex user problems into agent-driven workflows---Required Senior-Level Qualifications· 6–10+ years of professional software engineering experience· Demonstrated experience building and maintaining production agent systems with real users· Deep hands-on expertise with:o Lang Graph (graph-based orchestration, state persistence, supervisor/worker patterns)o Pydantic (advanced validation, custom types, schema evolution, backward compatibility)o Lang Smith (trace analysis, evaluation datasets, regression detection)· Strong Python expertise, including async patterns and concurrency· Experience designing idempotent and retry-safe agent actions· Proven ability to identify and mitigate LLM failure modes (hallucinations, tool misuse, infinite loops, degraded reasoning)· Experience implementing guardrails, policies, and safety layers in agent systems· Deep understanding of observability for AI systems beyond traditional logging---Advanced / Differentiating Skills· Experience designing multi-agent systems with delegation, arbitration, and supervision· Expertise in agent evaluation frameworks and continuous quality measurement· Experience with RAG architectures tuned for agent decision-making (not just QA)· Familiarity with semantic caching, prompt/version management, and deterministic replays· Experience deploying agent systems in high-availability cloud environments· Ability to mentor other engineers on agent architecture and operational best practices---What Sets This Role Apart· This role owns agent architecture, not just implementation· Success is measured by reliability and trust, not novelty· You will influence how agents are built across the organization· The problems are ambiguous, high-impact, and user-facing---Ideal Candidate Profile· Has shipped at least one agent-driven product to real users· Thinks in terms of systems, contracts, and failure modes· Comfortable saying "no" to unsafe or brittle agent designs· Treats agents as production software, not experiments

About Us: The CSI Companies is a leading staffing and recruiting firm, providing organizations with highly skilled professionals since 1994. We have been recognized as a "Best of Staffing" award winner for over a decade, and we pride ourselves on delivering exceptional talent to industry-leading clients. For consideration, please submit your resume with relevant experience. Only candidates selected for interviews will be contacted.

Benefits: Offered: Weekly pay Medical, dental, and vision coverage Voluntary Life and AD&D coverage Paid Training Opportunity for advancement upon performance and availability

Type: Contract

Category: Staff Augmentation

Date Posted: 2026-04-08