Job Description

What's the role?

We’re looking for an Agentic AI Data Scientist to drive learning, evaluation, and optimization of agentic, workflow-centric AI systems across logistics, fleet, and mobility solutions.

This role focuses on transforming real-world agent behavior into structured learning signals. You will collaborate closely with Agentic Domain Architects, product, and engineering teams to extract actionable insights from agent memory, user interactions, workflow outcomes, and human corrections, and convert them into training data, evaluation frameworks, and continuous optimization strategies.

This is a hands-on, applied data science role embedded in the product team with strong ownership across the full learning loop: signal → dataset → evaluation → improvement → production validation

Key Responsibilities

• Design and implement methods to extract high-quality learning signals from agent memory (short-term, episodic, long-term), user interactions and feedback, workflow outcomes, human corrections, overrides, and escalations.
• Partner with AI Domain Architects to ensure signals are semantically grounded in domain workflows, attributable to agent skills, decisions, and context.
• Distinguish between noise, bias, and learning-relevant signals in real production data.

• Transform raw agentic signals into high-quality training datasets
• Define labeling strategies that leverage implicit user behavior, outcome-based signals, Human-in-the-loop corrections.
• Establish data validation, versioning, and quality controls specific to agentic workflows.

• Design domain-specific evaluation frameworks that go beyond generic model benchmarks.
• Define and operationalize metrics for decision quality, goal completion, workflow efficiency, error handling and robustness.
• Partner with product and architecture to ensure metrics reflect business and operational value, not just model accuracy.

• Develop and apply optimization strategies such as fine-tuning, reward modeling and preference optimization for improving logistics process outcomes.
• Influence grounding, prompting, memory, and orchestration strategies using empirical evidence.
• Support experimentation frameworks (A/B testing, shadow deployments) for optimizing agentic behavior.

What Success Looks Like

• Agent behavior measurably improves based on real usage data
• Learning signals are reliable, interpretable, and action-oriented
• Evaluation reflects domain outcomes, not abstract benchmarks.
• Human corrections actively reduce future errors
• The organization builds a scalable learning infrastructure for agentic AI

Who are you?

You are a data scientist with strong applied instincts, excited about learning from real-world system behavior rather than curated benchmarks. You are comfortable operating in ambiguity, designing new evaluation and learning patterns for agentic AI systems rather than classical model-centric pipelines.

You enjoy working at the intersection of AI behavior, data, and product outcomes, and you care deeply about making customer data feedback loops trustworthy, measurable, and impactful.

Required Experience & Skills

• Bachelor’s or Master’s degree in Data Science, Computer Science, Machine Learning, Statistics, or a related field.
• 5+ years of experience in applied data science or machine learning, with exposure to production systems.
• Strong experience working with behavioral data, interaction logs, and outcome-based signals.
• Strong skills in SQL, Python (pandas, numpy, matplotlib, seaborn, or similar libraries), and data engineering (Spark, ETL, Kafka).
• Hands-on experience with: LLMs and/or agentic systems, model evaluation and experimentation, dataset curation and feedback loops
• Solid understanding of learning from human feedback, reward modeling or preference learning, trade-offs between deterministic logic and probabilistic models
• Strong analytical thinking with ability to translate data insights into product improvements
• Excellent communication skills—able to explain technical findings to diverse audiences
• Self-motivated with ability to proactively identify problems and propose solutions
• Experience working in fast-paced, cross-functional environments.
• Fluency in spoken and written English.

Nice-to-have:

• Background in mapping or GIS
• Domain knowledge in logistics processes, route optimization, routing algorithms, operations research

What we offer:

• Work on on globally deployed automotive navigation and mapping products
• Collaboration with an experienced, international engineering organization
• Truly international team of fantastic & talented people from 60+ countries worldwide, working from strong tech hubs located in Europe, the US, and Asia, along with multiple smaller locations
• Flexible working hours & a flexible approach to office vs. remote work
• Medical coverage for you and your family
• Yearly variable bonus
• This role is eligible for the Creative Tax Incentive scheme in Poland (KUP)
• Option to work on a B2B contract (please note: benefits, bonus and KUP do not apply in this case)

Life at HERE in Poland comes with a competitive total rewards package designed to support your health, wellbeing, and performance. This includes a base salary, a Short-Term Incentive (STI) bonus (percentage based on role), a creative tax advantage for eligible positions, private medical care (including dental), life insurance, a meal allowance, vision reimbursement, a remote work allowance (if applicable), access to MyBenefit and Multisport programs, and various wellbeing initiatives. Paid time off, sick leave, and parental leave are provided in accordance with the Polish Labor Code.

Who are we?

HERE Technologies is a location data and technology platform company. We empower our customers to achieve better outcomes – from helping a city manage its infrastructure or a business optimize its assets to guiding drivers to their destination safely.

At HERE we take it upon ourselves to be the change we wish to see. We create solutions that fuel innovation, provide opportunity and foster inclusion to improve people’s lives. If you are inspired by an open world and driven to create positive change, join us. Learn more about us on our YouTube Channel.

Change is HERE. Apply Now.

#LI-NK3 #LI-HYBRID

As part of HERE Technologies employment process, candidates will be required to successfully complete a pre-employment screening process. This offer and any related claims are subject to the successful completion of a pre-employment screening. This will involve employment, education, and criminal verification if applicable.

HERE is an equal opportunity employer. We evaluate qualified applicants without regard to race, color, age, gender identity, sexual orientation, marital status, parental status, religion, sex, national origin, disability, veteran status, and other legally protected characteristics.