AI Engineer

01 ABOUT NAVLOGIC AI

NavLogic AI was founded by veterans of the supply chain and logistics industry who have spent decades navigating the complexity of global freight, warehouse operations, last-mile delivery, and demand planning. Frustrated by legacy software that lagged years behind real-world challenges, our founders set out to build the AI-native platform the industry actually deserves.

We sit at the intersection of deep operational expertise and cutting-edge AI — applying large language models, computer vision, and predictive analytics to the messy, high-stakes world of physical goods movement. Our customers range from regional carriers to Fortune 500 manufacturers, all looking for intelligence that works as hard as they do.

02 THE OPPORTUNITY

We are looking for an AI Engineer who treats production-grade AI as an engineering discipline — not a research demo. You will design, build, and operate the models and AI systems that power NavLogic AI: LLM-based copilots for ops teams, computer-vision models that watch warehouse docks, forecasting and routing systems that move real freight.

You will own model lifecycle end to end: data, evaluation, training or fine-tuning, deployment, monitoring, and the inevitable 2 a.m. drift investigation. You partner closely with Forward Deployed Engineers to harden your work against real customer data, and with Product to ship features that customers actually feel. If you are equally fluent reading a PyTorch traceback and a Lamb-Howell delivery curve, this role was designed for you.

03 WHAT YOU WILL DO

•     Production AI Systems Design, train, evaluate, and ship AI features into production — LLM-based ops copilots, RAG over operational documents, computer-vision models for warehouse and yard, demand forecasting, and route/load optimization.

•     Model Architecture & Selection Make principled trade-offs between frontier API models, open-weights models, and bespoke fine-tunes. Justify choices with cost, latency, accuracy, and customer-data-residency in mind.

•     Evaluation & Observability Build the eval harnesses, golden datasets, and online metrics that tell us — quantitatively — whether a model change is a win. Drive a culture of “no eval, no merge.”

•     Inference Infrastructure Own latency, throughput, cost, and reliability of model serving. Quantization, batching, caching, autoscaling — whatever it takes to keep inference under SLO.

•     Agentic Workflows Build multi-step agentic systems that can plan, call tools, and recover from failure inside real logistics workflows (dispatch, exception handling, document processing).

•     Data & Feedback Loops Partner with FDEs and customers to instrument feedback, capture labels, and close the loop between production behavior and the next model iteration.

•     Research-to-Product Track frontier research; rapidly prototype and de-risk what is genuinely useful for logistics, and ruthlessly discard what is not.

•     Responsible AI in High-Stakes Ops Design for explainability, robustness, and graceful degradation in environments where “the model said so” isn’t a good answer when a truck is waiting.

04 WHAT WE’RE LOOKING FOR

Required Qualifications

•     5 years building and shipping ML/AI systems in production — not just notebooks, not just research.

•     Strong Python; deep familiarity with at least one of PyTorch, JAX, or TensorFlow; fluency with the modern LLM stack (prompting, RAG, fine-tuning, evals, agents).

•     Experience operating models in production: serving, autoscaling, monitoring, on-call, and the unglamorous work of keeping things up.

•     Experience designing rigorous evaluation pipelines — offline and online — and using them to drive model decisions.

•     Strong software engineering fundamentals: testing, CI/CD, code review, system design. You write code others can build on.

•     AI-native mindset: you use LLMs, code-gen agents, and AI-assisted research as core tooling in your daily workflow.

•     Comfortable with the ambiguity of an early-stage startup — capable of taking a half-formed problem and shaping it into a shipped feature.

•     Bachelor’s in Computer Science, Machine Learning, Statistics, or related field (or equivalent demonstrable experience).

Preferred / Nice-to-Have

•     Graduate degree (MS or PhD) in ML, CS, OR, Statistics, or a related discipline — and the wisdom to know when the academic answer is the wrong one for production.

•     Hands-on computer vision experience (object detection, segmentation, OCR, video) — bonus for warehouse, dock, or yard applications.

•     Time-series forecasting or operations-research experience — vehicle routing, MILP, network flow, demand sensing.

•     Experience with vector databases (pgvector, Pinecone, Weaviate), modern eval tooling (LangSmith, Braintrust, Inspect), and LLM observability (Langfuse, Arize, Helicone).

•     Logistics, supply-chain, freight, or industrial-IoT domain background — or genuine curiosity about it.

•     Open-source contributions, published research, or a portfolio that shows how you think.

•     Veteran or DOD-AI/logistics background.

05 THE AI-NATIVE EXPECTATION

At NavLogic AI, “familiar with AI tools” is table stakes — not a differentiator.

▸ You ship faster because you build with AI — agents, code-gen, eval-gen — and you understand their failure modes deeply.

▸ You experiment with new models, papers, and tooling before you’re asked to, and bring back what’s real.

▸ You can defend or kill a model decision with data — not vibes.

▸ You can explain, in plain English, why a forecasting model is wrong on Tuesdays — and what you’re going to do about it.

▸ You take seriously the ethical and reliability dimensions of AI in physical supply chains: bias, explainability, and the cost of a wrong answer at scale.

06 WHAT WE OFFER

•     Competitive Compensation Base salary performance bonus equity — benchmarked to Series A / B market rates.

•     Flexible Work Remote-first with optional hub access; async-friendly culture that respects your time zone.

•     AI Tooling Budget Dedicated annual budget for AI subscriptions, tools, and experimentation — we put our money where our values are.

•     Learning & Development Conference attendance, online courses, and a quarterly book/resource allowance; we invest heavily in keeping you sharp.

•     Mission-Driven Team Work alongside supply-chain veterans and AI engineers who have operated in the field and built production systems — no hand-wavy hype here.

•     Impact from Day One As an early hire you will shape the function, the product, and the culture — your fingerprints will be on everything.

•     Health & Wellness Comprehensive medical, dental, and vision; mental-health support; and generous PTO including military/veteran observances.

07 HOW TO APPLY

We review applications on a rolling basis. To stand out, we encourage you to include a short note (3–5 sentences) on a production AI system you have shipped — what it does, how you evaluated it, and what surprised you in production.

Applications without this note are still welcome; it simply helps us move faster.

Apply at: careers@navlogic.ai  |  Subject: AI Eng – [Your Name]