About the role
Defines and drives the technical direction of Tekion's Orbit platform. Operates at the intersection of deep hands-on engineering and cross-team architectural leadership. Brings 8-12 years of software engineering experience anchored in Java, with proven expertise in agentic orchestration, LLM pipeline design, and production-grade AI systems. Serves as a technical authority, influencing platform strategy, raising the engineering bar, and enabling multiple teams to move faster and build better.
Responsibilities:
- Own and drive the end-to-end technical architecture of AI agent systems, including multi-agent orchestration, LLM integration, vector and graph data layers, and durable workflow execution.
- Define platform-level patterns and reusable frameworks for agentic development using LangChain, LangGraph, and LangFlow, enabling other teams to build reliably on top of shared foundations.
- Lead the design and adoption of Qdrant-based semantic search and retrieval pipelines, and establish standards for knowledge graph modelling using Neo4j across platform teams.
- Architect fault-tolerant, long-running agent workflows using Temporal, defining reliability and retry patterns that serve as the organizational standard.
- Establish and own the observability strategy for AI workloads covering LLM call tracing, token cost attribution, latency SLOs, and agent decision transparency.
- Partner with engineering managers, product managers, and architects to translate business goals into multi-quarter technical roadmaps.
- Identify and resolve systemic technical risks, cross-team dependencies, and architectural gaps before they become production incidents.
- Conduct architecture reviews, set coding and design standards, and drive consistency across teams building on the AI platform.
- Mentor senior and mid-level engineers; invest in technical growth across the organization through design reviews, documentation, and knowledge sharing.
- Represent engineering in cross-functional forums; communicate technical strategy and trade-offs clearly to CTO, VP, and product leadership.
- Champion quality at a platform level to define test strategy, enforce contract testing between services, and drive observability-driven development practices.
Requirements:
- 9-12 years of professional software engineering experience with Java as the primary language.
- Demonstrated experience architecting and operating LLM-based agentic systems in production using LangChain, LangGraph, or LangFlow.
- Deep expertise with Qdrant or equivalent vector databases for large-scale semantic search and retrieval-augmented generation (RAG) pipelines.
- Strong experience with Neo4j or graph databases for complex domain modelling, knowledge graphs, and relationship-driven query patterns.
- Proven experience with Temporal or a comparable durable execution platform (Conductor, Cadence) for orchestrating fault-tolerant, long-running workflows at scale.
- Expert-level understanding of multi-agent orchestration: planning, tool use, ReAct and reflection patterns, memory architectures, and agent evaluation.
- Strong observability engineering background: OpenTelemetry, distributed tracing, metrics pipelines, LLM-specific telemetry, and SLO/SLA management.
- Track record of defining and enforcing engineering quality standards, test architecture, coverage strategy, code review culture, and zero-defect production mindset.
- Bachelor's or master's degree in computer science, engineering, or a related technical field.
- Excellent written and verbal communication; able to influence without authority across engineering, product, and executive stakeholders.
Preferred:
- Experience building internal developer platforms or shared engineering frameworks adopted by multiple product teams.
- Familiarity with LLM cost governance token budgeting, model tiering, context window optimization, and inference cost attribution at scale.
- Exposure to automotive retail technology, dealer management systems, or enterprise SaaS platforms with high regulatory or reliability requirements.
- Contributions to open-source AI or agent orchestration tooling, or published technical writing in the agentic systems space.
- Experience with security hardening for LLM systems: prompt injection defenses, output validation, data residency, and PII handling in AI pipelines.