Remote Data Architect for AI-Driven Platform Rebuild

UtiliSave LLC; New York (NY); Remote; USD 120,000 - 190,000; Full time

Benefits offered by this job

Annual bonus

401(k) with company match

Health insurance

Dental insurance

Vision insurance

Flexible spending accounts

Paid time off

11 paid holidays

Job summary

UtiliSave LLC is seeking a data engineering leader to own the data model across SQL Server and Dataverse, and to design the target model for Dynamics 365 CRM before migration. You will clean up and migrate decades of data, create robust data flows and integrations across SharePoint, finance systems, and our AI platform, while maintaining legacy systems during the transition.

You will implement strong data security, oversee Power BI semantic models, and ensure data quality with dedupe rules and a

Qualifications

• 5+ years in data engineering, data architecture or database development.
• Strong hands-on SQL Server: T-SQL, schema design and performance tuning.
• Experience with Dataverse or the Dynamics 365 data model: tables, relationships, keys and security roles.
• Built and supported production integrations using REST APIs, Azure Data Factory, Azure Functions, Power Automate or similar.
• Data modeling and data quality in practice: you're migrated a legacy database, wrote matching and dedupe rules, and set up a single source of truth.
• Data security experience: role-based access and permissions across databases, Microsoft 365, SharePoint and Entra ID.
• Power BI data modeling (semantic models, DAX), beyond building visuals.
• Working knowledge of Microsoft 365 and SharePoint: libraries, metadata, permissions and the Graph API.
• Demonstrated, hands-on use of AI-assisted development tools in real projects, not just familiarity. Claude Code preferred; Copilot, Cursor or similar accepted.
• Able to plan data flows and explain them clearly, in writing and out loud, to developers and non-technical colleagues.
• Comfortable working independently in a small, fully remote team.

Responsibilities

• Own the company's data model across SQL Server and Dataverse, and define the target model for our Dynamics 365 CRM before data is migrated into it.
• Plan and run data cleanup and migration: one trusted ID for each account, property, utility and contact, clear matching and dedupe rules, and intake checks that stop bad data from coming back.
• Plan, design and build data flows and integrations between SQL Server, Dataverse, SharePoint, finance systems and our AI platform, using APIs, Azure services and Power Automate.
• Protect the data: design and maintain permissions for staff, clients, auditors and AI tools across SQL Server, Dataverse, SharePoint and Entra ID.
• Own the reporting data layer, including Power BI semantic models and the history tracking we need to measure cycle times and outcomes.
• Support and maintain our existing applications and databases during the transition, so the business keeps running while the new platform comes online.
• Bring production integrations under standard controls: source control, code review, test environments and documentation.
• Use AI-assisted tools (Claude Code, GitHub Copilot, Cursor or similar) as a normal part of your work for SQL, pipelines, testing and documentation.
• Track all work in JIRA, with estimation and task breakdown before development begins, and write documentation as part of delivery.

Skills

Data modeling

Data architecture

Data quality

Plan data flows

Tools

SQL Server

Dataverse

Power BI

Azure Data Factory

Azure Functions

Power Automate

SharePoint

REST APIs

Job description

UtiliSave LLC is seeking a data engineering leader to own the data model across SQL Server and Dataverse, and to design the target model for Dynamics 365 CRM before migration. You will clean up and migrate decades of data, create robust data flows and integrations across SharePoint, finance systems, and our AI platform, while maintaining legacy systems during the transition.

You will implement strong data security, oversee Power BI semantic models, and ensure data quality with dedupe rules and a