Note: The job is a remote job and is open to candidates in USA. Dynatron Software, Inc. provides intelligent SaaS solutions, analytics, automation, and AI-powered workflows for automotive dealership service departments. The Lead Applied AI/ML Data Scientist will serve as a technical authority, owning production machine learning capabilities, shaping AI product strategy, developing generative AI and agentic systems, and establishing rigorous evaluation and production lifecycle standards.
Responsibilities
- Own modeling approaches for Dynatron's core classification, prediction, and other applied machine learning use cases
- Design features and modeling strategies for complex, messy, real-world automotive data
- Establish rigorous approaches to class imbalance, validation, experimentation, and model evaluation
- Continuously improve models based on production performance, changing data, and customer outcomes
- Raise the standard for how applied machine learning is developed, evaluated, documented, and shipped across the organization
- Design and build AI/ML capabilities from initial problem definition through production release
- Translate customer and product problems into appropriate modeling approaches rather than beginning with a predetermined technology
- Build solutions that balance model quality, scalability, explainability, latency, cost, and maintainability
- Remain engaged after launch to understand real-world performance and improve capabilities based on production evidence
- Serve as a technical authority on the feasibility of proposed AI capabilities
- Partner with Product leadership to evaluate opportunities before significant engineering investment is made
- Clearly articulate what is technically achievable, what requires additional data or sequencing, and what is unlikely to deliver the intended result
- Recommend alternative approaches when AI isn't the appropriate solution
- Help prioritize opportunities based on customer value, technical feasibility, data readiness, and implementation complexity
- Develop production capabilities using LLMs and modern agentic frameworks where they provide meaningful product value
- Design retrieval architectures, tool-use patterns, and other approaches for grounding AI systems in Dynatron's proprietary data
- Evaluate and adapt foundation models for domain-specific applications, including fine-tuning where appropriate
- Establish appropriate controls around quality, latency, token usage, and cost per interaction
- Stay current with emerging AI capabilities while applying disciplined judgment about where they belong in production
- Establish rigorous evaluation methodologies for traditional ML and non-deterministic generative AI systems
- Define appropriate offline and production metrics for individual use cases
- Design evaluation frameworks that measure accuracy, reliability, business usefulness, and other relevant quality dimensions
- Monitor production performance and use real-world results to guide model improvement
- Help establish consistent standards for determining when an AI capability is ready for customers
- Work closely with Engineering and MLOps/DevOps partners to establish production-readiness criteria
- Define requirements for deployment, monitoring, retraining, and model lifecycle management
- Ensure appropriate handoffs without treating productionization as someone else's problem
- Collaborate across Data Engineering, Product, and Engineering to ensure AI solutions have the data and infrastructure required to perform reliably
Skills
- 10+ years of experience in Data Science, Applied Machine Learning, or a closely related discipline
- Deep expertise in traditional machine learning, including classification, feature engineering, class imbalance, and model evaluation
- Significant experience working with complex, imperfect real-world datasets rather than exclusively curated research data
- Strong understanding of experimental design and how to determine whether a model is actually improving an outcome
- Demonstrated experience shipping AI/ML capabilities into commercial products used by real customers
- Ability to speak specifically about systems you've built, their scale, how they performed after release, and what you changed based on production evidence
- Experience supporting and improving models throughout their production lifecycle
- Strong understanding of the differences between building a successful prototype and operating a successful AI product
- Production experience building with LLMs and agentic frameworks
- Experience designing retrieval architectures and grounded AI applications
- Strong understanding of evaluation methodologies for non-deterministic systems
- Experience managing quality, latency, token consumption, and cost per interaction in production
- Experience fine-tuning or otherwise adapting transformer models for domain-specific use cases
- Demonstrated experience scoping AI initiatives directly with Product Managers and business stakeholders
- Track record of identifying technically weak or commercially impractical AI concepts and influencing stakeholders toward better solutions
- Ability to translate business problems into modeling problems—and recognize when the underlying problem doesn't require AI
- Strong customer orientation with curiosity about the business problem behind the requested capability
- Expert-level Python and strong SQL skills
- Comfortable working directly with large datasets in cloud data warehouse environments
- Experience collaborating within modern cloud-based data and ML ecosystems
- Strong understanding of the data requirements and dependencies necessary to support production AI
- Ability to communicate sophisticated AI concepts, limitations, and trade-offs clearly to non-technical stakeholders
- Strong influence skills and confidence challenging assumptions constructively
- Ability to establish technical standards and raise the quality of work without direct people-management authority
- Strong documentation habits and commitment to making technical decisions understandable and reproducible
- Master's degree in Computer Science, Data Science, Statistics, Mathematics, Engineering, or another quantitative discipline, or equivalent practical experience
- Experience working with automotive, dealership, or Fixed Operations data
- Experience in another domain involving complex operational records, industry-specific taxonomies, or similarly challenging datasets
- Experience with cloud-managed AI/ML services and modern production model lifecycle practices
- Experience designing, managing, or governing large-scale expert labeling programs
- Experience working with proprietary datasets as a foundation for differentiated AI products
Benefits
- Comprehensive health, dental, and vision insurance
- Equity participation through Dynatron's Equity Incentive Plan
- 401(k) with competitive company match
- Flexible vacation policy and 11 paid company holidays
- Employer-paid short- and long-term disability and life insurance
- Home office setup support
- Remote-first working environment
- Ongoing professional development opportunities
Company Overview