Global Head of Data Ingestion

Company:
Marsh – Corporate

Description:

We are seeking a Global Head of Data Ingestion to join our Chief Innovation & Data Office team s role will be based in New York, Boston, or London. This is a hybrid role that has a requirement of working at least three days a week in the office.

The Global Head of Data Ingestion is a senior leadership role within the Chief Innovation & Data Office, responsible for designing, executing, and continuously improving the enterprise strategy for acquiring, validating, and delivering high‑quality data s leader will define how data, both structured and unstructured and from internal systems and external partners alike, flows into a golden‑source data environment ready for analytics, AI, and commercial use, as well as into downstream systems.

The role carries both strategic authority over the ingestion architecture and operational accountability for cost, speed, and accuracy.

The position encompasses three interrelated mandates. First, helping craft the domain‑led strategy for what data to ingest and in what priority order, working closely with business leaders to connect those efforts and investments to commercial value. Second, owning the technical architecture and tooling that makes ingestion fast, repeatable, and scalable; that includes oversight of engineering teams and vendor partnerships that build and maintain bespoke ingestion products.

Third, leading a centralized human‑in‑the‑loop Capability Center that provides final data validation where automated methods alone are not yet sufficient, with an explicit mandate to shrink that manual footprint over time as models and processes mature.

This is neither a pure strategy role, nor a pure operations role. We are looking for someone who can articulate a compelling vision for enterprise data ingestion, hold genuine technical authority over how tooling is built inside of our enterprise architectures, and deliver hard financial results through consolidation, automation, and industrialization. The successful candidate will have a builder’s instinct, a strategist’s perspective, and the pragmatism to know that some of the most valuable data in insurance still requires a human to verify it.

• Define and maintain a prioritized, domain‑based ingestion strategy that ties every ingestion workstream to a specific commercial or operational outcome.
• Partner with business leaders across the enterprise to identify which data elements and sources drive the highest‑value outcomes and prioritize their ingestion over lower‑value noise.
• Break down line‑of‑business silos to ensure data is ingested in a way that supports a unified client view across the enterprise, rather than fragmented, business‑unit‑specific narratives.
• Navigate the diversity of ingestion challenges across the enterprise: from extracting structured information out of PDF documents and emails, to remapping and harmonizing data that arrives structured but in incompatible formats from HR systems, claims platforms, or carrier feeds.

• Own the target‑state ingestion architecture: define the standards, patterns, and guardrails for how data enters the enterprise data environment (Databricks/MIDAS), whether through automated pipelines, API‑based integrations, or human‑assisted workflows. Where appropriate, define how ingested data is piped directly to downstream operational systems.
• Drive modernization from bespoke, fragile integration points toward scalable, repeatable ingestion patterns that reduce the marginal cost of onboarding new data domains.
• Bring deep understanding of how large language models and AI can be applied to ingestion problems: extracting structured fields from unstructured documents, classifying and routing content, and generating deterministic code to automate schema mapping and data transformation between systems.
• Maintain hands‑on technical credibility: you will not write production code, but you must be able to design solutions, evaluate engineering proposals on their merits, and hold technology partners to a high bar on architectural quality.
• Refuse short‑term…