Senior AI Software Engineer

Jackson Dawson•Dearborn, MI

16h•Hybrid

About The Position

Livium is seeking an experienced and entrepreneurial Senior Software Engineer (AI/ML) to join our growing engineering team building products such as TrueTrack, a financial reporting and analytics platform for powersports and automotive dealerships. You will work with software developers to guide AI implementation strategies, build data pipelines, and actively explore new AI applications. We are looking for engineers who are passionate about applying AI/ML to solve real-world business problems, enjoy working with large datasets, and thrive in a collaborative startup environment where they can make a direct impact. About Livium Livium is a technology company that emerged from Jackson Dawson, bringing 40+ years of business expertise to deliver innovative technology solutions. At Livium, we deliver practical solutions that help businesses gain actionable insights from their data, enabling better decisions, improved operational efficiency, and the ability to focus on what matters most.

Requirements

3–6 years of software engineering experience, with at least 2 years in applied ML or AI systems
Expertise in Python and popular ML libraries (TensorFlow, PyTorch)
Familiarity with Agentic Frameworks (LangGraph, LangChain, CrewAI)
Hands-on experience building with LLM APIs (OpenAI, Anthropic, Bedrock, etc.) in production.
Practical knowledge of evaluation methodologies for LLM systems.
Familiarity with gradient boosting methods (XGBoost, LightGBM, scikit-learn) applied to structured/tabular data
Strong understanding of data structures, algorithms, and distributed computing

Nice To Haves

Master's degree in Computer Science, Statistics, or related field with ML focus
Experience with cloud platforms (AWS, Azure, GCP) for AI/ML
Experience with RAG architectures and vector databases
Background in financial analytics or SaaS B2B products
LLMOps Experience (LangFuse, Evals, Hallucination Avoidance, Guardrails)
MLOps Experience (ETL Pipelines, Model Training, Deployment, Monitoring, Rollbacks)
Exposure to prompt optimization, fine-tuning workflows, and structured outputs
Experience in a startup or scale-up environment where you wear multiple hats

Responsibilities

Design, build, and deploy secure multi-agent systems using LangGraph and similar orchestration frameworks
Implement production guardrails including output validation, content filtering, and fallback strategies
Develop and maintain LLM evaluation frameworks (evals) to measure response quality, hallucination rates, and task accuracy
Instrument AI systems with observability tooling (traces, latency, token cost monitoring)
Stay current with the AI ecosystem and evaluate emerging frameworks, models, and patterns for relevance to the product
Train, validate, and deploy tabular ML models (gradient boosting, XGBoost/LightGBM) on structured transactional data
Develop models for pricing optimization, inventory forecasting, and related predictive tasks for different industries.
Design feature engineering pipelines and manage model versioning and retraining schedules
Select appropriate modeling approaches for each problem, from statistical baselines to LLMs, based on data characteristics and business constraints
Write production-quality Python services; own the full lifecycle from design to deployment
Integrate AI/ML capabilities into existing platforms via clean, well-documented APIs
Collaborate with backend engineers on data pipelines and infrastructure requirements
Participate in code review and maintain engineering quality standards