Senior AI Software Engineer (Architect of Autonomous Systems)

Ron Turley Associates LLC•Glendale, AZ

3d•$180,000 - $220,000•Hybrid

About The Position

RTA is looking for a Senior AI Software Engineer who doesn't just use AI — you build the systems that run on their own. This is a role for someone who thinks in agents, reasons in pipelines, and has a track record of shipping AI systems that actually work in production — not just in demos. You'll be a core technical contributor on our AI platform efforts, helping Fleet360 evolve from a traditional FMIS into an intelligent, self-improving system that thinks alongside our customers. You will report directly to the VP of Innovation and work closely with our product and engineering leadership to design and deliver agentic AI features, self-improving systems, and intelligent automation that sets RTA apart in the market. This role isn't about writing wrappers around LLMs. It's about building the kind of AI systems that detect anomalies, recover from failure, orchestrate multi-step reasoning, and make fleets smarter — without someone holding their hand. We also believe deeply that the best AI engineers use AI heavily in their own workflows. We're looking for someone who has genuinely rethought how they work because of these tools — not someone who just knows how to talk about them. If you're the kind of engineer who builds things that run themselves, ships AI that scales, and uses every tool at your disposal to do it faster and better — we'd love to talk.

Requirements

5+ years of professional software engineering experience
2+ years of hands-on experience building production AI/ML systems — not just prototypes
Deep experience designing and shipping agentic AI systems using orchestration frameworks, tool-calling, and multi-step reasoning pipelines
Proven experience fine-tuning LLMs on domain-specific data, including dataset preparation, training runs, and rigorous evaluation
Experience building self-improving or self-monitoring systems — systems that observe their own outputs, measure quality over time, and automatically adjust based on production signals
Strong software engineering fundamentals: clean code, system design, testing, and production reliability
A genuine track record of using AI heavily in your own day-to-day work — not just building it for others
Experience with vector databases, embeddings, RAG pipelines, or semantic search in production environments
Strong communication skills and ability to work across engineering, product, and business stakeholders

Responsibilities

Build Agentic AI Systems
Design and ship multi-step AI agents that can reason, plan, and take action across Fleet360 data and workflows
Build agent orchestration frameworks using patterns like ReAct, tool-calling, and multi-agent coordination
Architect reliable agent pipelines with robust error handling, fallback strategies, and human-in-the-loop checkpoints where they matter
Use orchestration frameworks (CrewAI, LangGraph etc.) to build, manage, and scale Agentic AI systems
Design agents that don't just respond to requests, but proactively surface insights and take autonomous action on behalf of fleet operators
Build Self-Improving Systems
Design systems that observe their own outputs, measure quality over time, and automatically adjust based on what they learn from production
Build feedback loops that close themselves — where model behavior, agent decisions, and pipeline outputs feed back into improved performance without manual intervention
Implement continuous evaluation pipelines that track model and system quality in production, detect degradation early, and trigger retraining or recalibration automatically
Design data flywheels where every user interaction makes the system smarter — capturing signal, labeling at scale, and feeding it back into fine-tuning and prompt optimization
Architect systems with self-assessment built in — agents that know when they're uncertain, recognize the boundaries of their own competence, and route accordingly
Fine-Tune and Adapt LLMs
Run structured fine-tuning experiments on domain-specific datasets to improve model performance for fleet and maintenance use cases
Design and manage evaluation pipelines that measure model quality reliably before and after changes
Make principled decisions about when to fine-tune, when to prompt-engineer, and when to reach for a different tool entirely
Maintain a clear-eyed view of model tradeoffs — latency, cost, accuracy, hallucination risk — and make decisions accordingly
Automate Your Own Work (and Ours)
Bring a genuine AI-native approach to your daily workflow — using agents, scripts, and AI tools to reduce repetitive work, accelerate iteration, and free up cognitive bandwidth for harder problems
Build internal tooling and automation that helps the product & engineering team move faster
Champion a culture where AI is a standard part of the engineering toolkit, not an occasional experiment
Share what you're learning — your teammates should get smarter about AI because you're on the team
Collaborate Across the Organization
Partner with Product, Engineering, and Customer Success to identify high-value AI opportunities in Fleet360
Translate fuzzy product goals into concrete AI system designs with honest assessments of feasibility and risk
Communicate clearly about what AI systems can and can't do — especially to non-technical stakeholders
Contribute to RTA's broader AI roadmap and help shape how intelligence gets woven into the platform over time