AI CoE Senior Data Engineer & Architect

Job Description

• About the team

QAD | Redzone is building an enterprise AI platform to support AI-enabled and agentic workflows across the

organisation. The AI Centre of Excellence (CoE) is accountable for turning that ambition into deployed,

measured, production workflows across priority business functions.

The CoE is small, senior and high-leverage. AI Engagement Specialists own the demand side with the

functions; the AI Architect and Senior AI Engineer build the platform and the agents; AI Operations owns the

run side. The Senior Data Engineer & Architect owns what every agent depends on: getting the right data to agents, current, governed and in a shape they can use.

• Why this role exists

Every AI use case in the CoE portfolio depends on data: from sales and customer systems, from support and

engineering knowledge sources, and from delivery and operational systems. Before an agent can use it, that

data has to be joined, current, governed and checked against what each user is allowed to see.

This role makes data a shared, reusable capability of the AI CoE rather than something rebuilt for each use

case. It owns the data layer at feasibility and development) stages and keeps the data running after go-live.

Role overview

The AI CoE Senior Data Engineer & Architect owns the data layer of the AI lifecycle. You will assess data

feasibility for each new use case, build and run the pipelines that feed the AI platform, model the semantic

layer that agents query instead of raw tables, and make sure every agent sees only the data its user is

allowed to see.This is a hands-on engineering role with production ownership, measured on agents getting correct, current and permitted data, and on each new use case reaching data-ready faster than the last.

• Where this role sits

Reports to Head of the AI Centre of Excellence

Works with Senior AI Engineer daily; AI Architect on reference architecture; AI

Engagement Specialist on data feasibility; AI Ops Lead on pipeline health

and cost; system integrators; data owners in each function; Security and

GRC

Owns Data feasibility assessments, AI data pipelines and connectors, semantic

models and metrics for AI use, retrieval-ready knowledge sources, data

access patterns for agents, and data freshness and quality targets

QAD | Redzone — AI CoE Data Engineer · AI Centre of Excellence

Accountable for Correct, current, governed data for every agent in production, measured

on freshness, quality, reuse of semantic models and time from PRD

approval to data-ready

Location India

• What you will own

• Data feasibility and readiness
• Assess the data behind every PRD before build is approved: sources, owners, volumes, quality, access

paths and gaps, written up as a data readiness view the stage gate can act on, with what is missing, stale

or not permitted and options to fix it.

Pipelines and integration

• Build and run the pipelines that feed the AI platform (on AWS today) and the warehouses agents query,

such as BigQuery or Redshift, with incremental and, where it matters, event-driven sync.

• Close freshness gaps: know how long a new record or document takes to reach an agent, set a target

with the business, and meet it. Handle deletions and permission changes, not just new records.

• Build and maintain connectors with the system integrators, so each new source is added once and

reused across use cases.

Semantic layer

• Model the business entities and metrics agents need in the semantic layer, so agents query governed

definitions agreed with data owners, consistent with what dashboards report, instead of writing SQL

against raw tables.

• Put guardrails on structured queries: read-only access, row limits, timeouts and cost limits for any agent

that queries data.

• Retrieval-ready knowledge

• Prepare unstructured sources for retrieval with the Senior AI Engineer: metadata, release and version

tags, access tags, and structure that keeps the source context intact; build and maintain the golden

datasets used to test retrieval and answer quality.

Data access, security and governance

• Make agents act as the user: user-scoped access, row and column-level security in the semantic layer,

and no over-privileged shared service accounts.

• Handle sensitive data properly with Security and GRC: classification, masking of personal and customer

data in tickets, residency rules for regulated regions, lineage and audit evidence.

Run and improve (L5)

• Own pipeline health in production: freshness and quality monitoring, alerts someone acts on, and fast

recovery when a source system changes its schema; with AI Ops, keep processing, storage and

embedding spend per use case in check.

• Turn each build into reusable data products so the next use case starts from shared models and

connectors, not from scratch.

Qualifications

• Essential
• 8+ years in data engineering, including at least 3 years building data for ML, analytics or Gen AI products

running in production.

• Strong SQL and Python, and pipeline orchestration in practice, for example Airflow, dbt, AWS Glue or

similar.

• Cloud data platforms: Azure , AWS (S3, Glue, Redshift, RDS) and/or Google BigQuery.

• Semantic and dimensional modelling: hands-on with a semantic or metrics layer such as Cube, dbt

Semantic Layer, LookML or Power BI semantic models.

• Incremental and event-driven ingestion: change data capture, webhooks or streaming, and handling

deletes and late-arriving data.

• Data for retrieval: document ingestion, metadata design and hybrid or vector search (for example

OpenSearch, Elasticsearch or pgvector), and enough understanding of embeddings to work with an AI

engineer.

• Data security in practice: role-based access, row and column-level security, handling personal data, and

cloud identity and access management.

• API integration with enterprise SaaS, such as CRM and Atlassian REST APIs.

• Data quality discipline: tests, freshness and quality monitoring, and service-level targets for data.

• Clear communication with data owners and business teams: able to explain what the data can and

cannot support, and why.

• Data depth

We are looking for someone who has run pipelines in production and lived with the consequences, and can

tell when an agent's wrong answer is really a data problem.

Strongly preferred

• Sales and customer data, support and engineering knowledge data, or ERP and manufacturing data;

Progress OpenEdge familiarity is a plus.

• Experience with an enterprise RAG or AI platform, and with exposing data to agents through APIs or

MCP tools.

• Working alongside system integrators and a central data team, and across time zones; enterprise B2B

SaaS experience.

•

• What this role is not
• Not a BI developer. Dashboards are owned by the functions and AI Ops; you make the data behind them

and the agents trustworthy.

• Not the owner of the enterprise data platform. You work with the central data team and reuse their

assets, and build what AI use cases need on top.

• Not an AI engineer. Agent logic, prompts and models sit with the Senior AI Engineer; you own the data

they depend on.

QAD | Redzone — AI CoE Data Engineer · AI Centre of Excellence

• Not a one-off integration contractor. You own the pipelines and semantic models in production, after

go-live.

What success looks like

First 3 months

• Data feasibility assessments completed for the priority use cases.

• A data readiness checklist built into the stage gates and used on every new use case.

• An agreed access pattern with Security for user-scoped data, and freshness measured for priority data

sources.

• Semantic layer version 1 live for the priority domain, used by production agents.

• Freshness and quality targets set and monitored for every production data source.

Additional Information

• Your health and well being are important to us at QAD. We provide programs that help you strike a healthy work-life balance.

…