Role: AI Product Tools & Frameworks Data Modeling Architect
Location: Noida & Hyderabad
Experience: 18-26 Years

Role Summary:
Own the canonical data model for the product — functional semantics and physical implementation — and validate that delivered schemas and pipelines conform to it. This is an architecture-and-governance role with sign-off authority, not a build-only IC role.

Key Responsibilities:
Define the canonical/logical data model and its business semantics; author and maintain the data dictionary
Design physical schemas, partitioning strategy, and performance characteristics on the target platform
Specify data contracts for adapters and CTO-org services; enforce contract versioning
Validate delivered schemas/pipelines against the model; hold sign-off authority — reject non-conforming delivery
Own data quality rules and lineage; partner with governance/compliance on PII classification

Must-Have Requirements:
14+ years in data modeling, with hands-on schema design (not just review) on at least one of: Databricks, Snowflake, Synapse, BigQuery




Demonstrated ownership of a canonical model consumed by 3+ downstream systems — ask for a specific example, not a claim
Robust dimensional (Kimball) and normalized (3NF) modeling; can explain SCD1/2/3 trade-offs unprompted
Can run an SME interview and produce a correct model — this should be a live exercise in the loop, not a resume line
Advanced SQL and query/performance tuning on the chosen platform
Working knowledge of PII identification and audit/balance/control practices for data
Must have build simple Agents to make Data Modeling Job more accurate and highly reliable

Preferred:
Governance/catalog tooling: Collibra, Alation, Purview
Compliance exposure: GDPR, CCPA
Modeling tools (Erwin, ER/Studio, dbdiagram, etc.)
dbt, MDM experience
Fundamentals of LLM-based agent tooling (ChatGPT/Claude/Gemini) applied to the domain — genuinely optional, not a screen-out

Domain Expertise (Relevant Business Vertical):
Demons

📌 AI Product Tools (Noida)
🏢 HCLTech
📍 Noida