EBay - Lead/Manager - AI Analytics Diagnostics

:

Analytics is undergoing a fundamental shift, moving from traditional data science and manual reporting toward AI-driven analytics product development. As business intelligence evolves into intelligent, AI-powered tooling, we are looking for someone eager to help lead this transformation.
This role sits at the critical intersection of business analytics, prototyping, and AI engineering. You will play a key part in reimagining our Business Intelligence ecosystem by stepping into the role of an AI Analytics Prototyper.
We are building toward a future where business and product teams can diagnose performance, uncover insights, and take action through AI-enabled experiences. If youre motivated to move beyond ad-hoc analysis into designing the blueprints for AI agents, mapping out deep business logic, and building functional prototypes that our developers will productionize, this is an opportunity to be at the forefront of shaping how eBay interacts with data in the AI era.
About the role :
Join us to act as the analytical engine behind our agentic AI systems. You will bridge the gap between complex business diagnostics and Agentic analytics product development. You will focus on deep-dive business analysis, creating the analytical logic, and building functional prototypes (e.g., via Python, Streamlit, or Jupyter). You ensure the AI system's answers are accurate, verifiable, logical, and easy for stakeholders to understand and act.
What youll do :

- Build AI analytics prototypes : Create functional prototypes using Python, Streamlit, SQL, and similar tools to show how AI agents should analyze data and generate business narratives.
- Drive Business Diagnostics : Deep dive into business performance (e.g., GMV variances, funnel conversion, traffic trends etc), uncover root-cause drivers, and translate this manual diagnostic logic into structured rules and workflows for AI agents.
- Validate & Evaluate AI Outputs : Define the queries, transformations, metrics, and validation logic needed to ensure AI-generated insights are accurate and verifiable.
- Bridge business and technical teams : Translate business needs into schemas, metadata, logic trees, and tool-call requirements for data product and AI development teams.
- Evaluate AI output quality : Test prototypes and model outputs for accuracy, relevance, statistical soundness, and hallucination risk.
- Partner Across Disciplines : Work with business leaders, data scientists, and AI developers to move prototypes toward scalable production systems.
Must-have :
- 8+ years in advanced data analytics, business performance diagnostics, or data science.
- Strong hands-on experience building functional data prototypes, automated workflows, or internal tooling (e.g., using Python, Pandas, Streamlit, Gradio, or Jupyter Notebooks).
- Deep expertise in SQL, data modeling, and dissecting complex business metrics (e.g., isolating mix vs. rate impacts, A/B testing, cohort analysis).
- Exceptional business acumenthe ability to look at a variance in a dashboard, manually diagnose the "why" behind it, and document that analytical thought process into a repeatable algorithm.
- Familiarity with LLM/agent integrations (understanding how structured tool calls, prompts, and retrieval-augmented generation work conceptually).
- Obsessive commitment to data accuracy, reproducibility, and explainability.
Nice to have :
- Background in e-commerce, marketplace analytics, or corporate finance/business performance tracking.
- Exposure to LLM prompt optimization, chaining, and human-in-the-loop evaluation.
- Experience partnering directly with software engineering or platform teams to hand off prototype code for productionization.
Success looks like :
- You successfully bridge the gap between executive business questions and Agentic Analytics development.
- Your prototypes and diagnostic blueprints enable the development team to build highly accurate, trustworthy AI agents that automate deep performance analysis.
- Because of your logic and testing, analysts and executives trust the AI's insights implicitly.