RGM Product Manager, AI Engineering

Coca-Cola; Georgia; On-site; USD 250,000 - 350,000; Full time

Job summary

The Coca-Cola Company is seeking a Senior Director of AI Engineering for the Global RGM Product. You will own the intelligence layer across the RGM suite, working in a cross-functional team to drive scalable, data-driven decisions at global scale.

You will lead the prompts, pipelines, retrieval, and evaluation processes, ensuring outputs are data-backed and trustworthy. This role operates across markets, requiring hands-on engineering leadership and strong collaboration with product management,

Qualifications

• Experience building data pipelines and analytics in Python.
• Familiarity with AI model integration and evaluation.
• Ability to specify prompts and QA stages clearly.

Responsibilities

• Own the prompt behind each pipeline node, versioned in source control.
• Extend the LangGraph orchestration pipeline in Python for data ingestion and KPI computation.
• Define and track evaluation metrics for every run and gate changes on improvements.
• Build retrieval over reference material so generated text quotes underlying data.
• Prototype new analyses on existing run data and promote to a feature after testing.

Skills

Python

AI/ML concepts

Data analysis

Tools

pandas

source control

Job description

Senior Director, RGM Product AI Engineering

About the Role

Revenue Growth Management (RGM) is one of The Coca-Cola Company's most important commercial capabilities, and not simply a pricing discipline. Coca-Cola positions RGM alongside brand building, innovation, and integrated execution as a core capability for topline growth, and our growth strategy names it a key commercial capability for determining where to play and how to win.

RGM connects consumer demand to the economics of the system. It determines what we sell, in which packages, at what price points, through which channels and customers, and with what promotional investment, all in service of balanced growth across transactions, volume, and price/mix. Done well, it finds incremental revenue pools and converts them into commercial action, improves the economics of every price, pack, assortment, and promotion decision, builds the affordability and premiumization ladders that bring more consumers into the portfolio, and gives Coca-Cola and its bottlers a common fact base for growing customer revenue and profit rather than simply negotiating price.

We are now reimagining RGM for the Agentic AI era, and we are building it ourselves. Our Global RGM Product is evolving from a suite of analytical solutions into an intelligent decision platform where AI agents interact directly with enterprise data, invoke advanced analytics and optimization models, run and compare scenarios, and hand commercial teams a recommendation with the reasoning attached. Work that once took an analyst a week (pulling the data, running the model, writing the story, defending the number) is being compressed into a conversation.

Few companies are attempting this at this scale, and fewer are still against decisions this consequential: real pricing, promotion, and portfolio moves, made across markets and bottlers. The work sits at the frontier of the CPG industry, spanning agent orchestration, retrieval over enterprise data, and the evaluation of non-deterministic and deterministic systems, and none of it is a research exercise. What we build ships to commercial teams who use it to make decisions the same week.

This is an opportunity to build, not just design, the next generation of commercial decision intelligence at Coca-Cola, and to establish the technical foundation for agentic RGM at global scale. The patterns set here will shape how one of the world's largest commercial systems makes its most important decisions. If you want your work in front of real users, at real scale, from the first release, this is that role.

Role Overview

As Product Manager of AI Engineering for the Global RGM Product, you will own the intelligence layer of the RGM suite within a persistent, cross-functional product team working alongside Product Management, Data Science, Engineering, UX, Architecture, and RGM experts. This is a Senior Director-level, hands-on engineering role, with technical direction provided by the RGM Product Tech Lead.

RGM turns price-optimization and promotion-simulation runs into decks and insights that RGM teams act on. A pipeline reads the run export, computes the KPIs in code, has the model write the narrative, and a QA step reviews the result before anyone sees it. You will own everything the model touches in that chain: the prompts, the pipeline steps, the retrieval, and the evaluation numbers that say whether an output can be trusted. One house rule sits above the rest-code computes every number, and the model only writes the words.

Working in a product model, you will remain engaged throughout the product lifecycle-from discovery and prototyping new analyses on existing run data through eval-gated release, deployment, measurement, and continuous improvement. The work is judged on evidence: a prompt or pipeline change ships when the quality numbers improve, not when it feels better. The role is global and operates around the clock, supporting markets and teams in every region.

What You Will Do for Us

• Own the prompt behind each pipeline node - validation, narration, slide planning, and QA review - versioned in source control alongside test inputs and expected outputs rather than maintained in a chat window.
• Extend the LangGraph orchestration pipeline in Python: ingest run exports, compute KPIs with pandas, call the model for narrative, and route QA retries when a generated slide fails review.
• Define and track the evaluation metrics for every run - the share of insights analysts rate gold or silver, the share hallucinated or missing a baseline, and volume coverage - and gate prompt changes on those numbers improving.
• Build retrieval over reference material, including methodology documentation, QA references, and past runs, so generated text quotes the underlying data instead of guessing.
• Prototype new analyses on existing run data, from new insight types to competitor-response summaries, promoting a prototype to a feature only when it has a te