Manager Account Analytics

Overview About Business Unit: Powerful predictors of the future - our Analytics team seamlessly blends data science and business intelligence to provide unparalleled foresight for our platform offerings. They turn insights into actionable results for our clients by collaborating with senior team members and staying ahead of market trends, powered by our data assets. By using new-age analytic tools and emerging technologies, the team cultivates a dynamic learning environment for curious minds to thrive. How You'll Make an Impact: The Manager, Account Analytics Development is a hands-on technical leader. You will personally design and build the most technically demanding parts of our AI capability, and you will hire and develop the engineer who works alongside you. This is a small, senior team — today a VP and a Lead. You will be the third member, and the first person you hire will report to you. We already run AI agents in production that generate analyst-grade SQL against our analytics environment. That foundation works. What comes next is harder, and it is the reason this role exists. What You'll Build: Interpretation. Systems that take the output of an analysis and produce business-relevant interpretation — what the numbers mean for this client, in this category, given what we know about their objectives. This is the distance between a query result and an insight, and closing it requires reasoning, not summarization. Client knowledge. A knowledge system that accumulates and organizes what the organization understands about each client: their business, their category, their measurement history, and Epsilon's relationship with them. It has to serve both people and agents, and it has to compound with use rather than decay into a stale wiki. Media investment planning. Analytical tooling that supports budget allocation and media investment planning — putting scenario and optimization capability directly in the hands of client teams. Click here to view how Epsilon transforms marketing with 1 View, 1 Vision and 1 Voice. Responsibilities Architecture and Systems Design Own the architecture of our AI capability: how agents are scoped, how they compose, and where the boundaries sit between tools, context, and model reasoning. Design evaluation before you design the solution. Define how we will know an agent is correct, how we detect regression in a non-deterministic system, and where human review belongs in the loop. Establish the engineering standards — source control, testing, documentation, deployment — that let a three-person team ship reliably today and a larger team ship reliably later. Hands-On Delivery Write production code. Prototype quickly to resolve technical uncertainty, then make the explicit call about what graduates into a supported system and what gets discarded. Take personal ownership of the components carrying the highest technical risk rather than delegating them. Team Leadership Hire, onboard, and develop your first direct report at senior engineer level. Set technical direction and raise the quality bar through design review and code review. Absorb technical scope currently carried by the VP and become a person the wider analytics organization brings its hardest problems to. Promote engineering discipline, documentation, code quality, and knowledge sharing across the team. Business Partnership Work directly with analytics leads and client teams, most of whom are US-based, to find where automation and AI create real leverage — and where they do not. Translate loosely specified business problems into system designs, and explain those designs credibly to non-technical partners. Balance experimentation against the reality that these tools enter client-facing workflows and have to keep working. Your work will be evaluated by the level of usage and adoption by the rest of the Analytics team and the ability of the tools you and the team develop to drive greater efficiency and insights Qualifications 8+ years building software or data products, including leading technical design on complex, ambiguous projects. Strong Python and advanced SQL. You are comfortable both in a substantial codebase and in an analytical data warehouse. Demonstrated experience designing and shipping LLM-based systems in production — agents, tool use, orchestration, structured outputs — not prototypes or prompt experimentation alone. Systems thinking. You can take an ambiguous capability such as “interpret this analysis” and decompose it into components, interfaces, and failure modes, then explain why you drew the boundaries where you did. A considered point of view on evaluating non-deterministic systems, and practical experience applying it. Working familiarity with marketing measurement — attribution, incrementality, media mix modeling, campaign performance — sufficient to judge whether an automated interpretation is analytically sound. You will not be building measurement models; you will be building systems that reason about their output. Experience mentoring or leading engineers, whether or not you have held a formal management title. Bachelor's degree in Computer Science, Engineering, Data Science, Mathematics, Statistics, or a related discipline. Additional Knowledge, Skills and Abilities Preferred Technical Experience Agentic development environments — Claude Code, Cursor, GitHub Copilot, or similar — used as a daily working tool rather than tried once. Databricks, Greenplum or PostgreSQL, Snowflake, BigQuery, or comparable analytical platforms. Cloud platforms (Azure, AWS, GCP), CI/CD pipelines, and disciplined code review practice. Building knowledge management or retrieval systems that stay accurate as they grow. Integrating applications with APIs, databases, and enterprise platforms. Marketing analytics, media, or adtech background. Working with data under client confidentiality and governance constraints. How You Work You are opinionated about design and you change your mind when the evidence