Senior Platform Architect — AWS & Databricks

About AbarVa
AbarVa is building AI-native enterprise platforms that help organizations make better technology, sourcing, data and AI decisions—and turn those decisions into measurable outcomes.
We are expanding our Data & AI delivery capability and are looking for a senior hands-on AWS / Databricks platform architect who can design, build and prove enterprise-grade Databricks environments in regulated organizations.
This is not a PowerPoint architecture role.
You will work directly with AbarVa leadership, senior client technology teams, Databricks, cloud infrastructure teams and our data-engineering leads to take an approved architecture from design into a secure, automated and operable platform.
What you will do
You will own the technical implementation of enterprise Databricks foundations on AWS, including:
AWS and Databricks platform architecture across Dev, Pre-Production and Production.
Databricks account, workspace and metastore configuration.
Unity Catalog architecture, catalogs, schemas, workspace bindings, grants, tags, masks and access policies.
AWS networking including VPCs, routing, PrivateLink, DNS, security groups and connectivity to enterprise data sources.
IAM roles, service principals, SSO/SCIM, secrets and least-privilege access.
S3 storage architecture and KMS/customer-managed encryption.
Serverless compute as the default pattern, with Classic compute where required.
Terraform/IaC for repeatable platform deployment rather than console-driven configuration.
CI/CD and Databricks Declarative Automation/Asset Bundle patterns.
Platform observability, audit logging, SIEM integration, monitoring and alerting.
Cost controls, tagging, budgets and FinOps.
Security, disaster recovery, backup/recovery and platform-operability controls.
Positive and negative readiness tests that prove infrastructure and access controls actually work.
Technical change packages for enterprise infrastructure teams, including exact configuration, security rationale, implementation steps and rollback.
Platform runbooks and knowledge transfer so the client's own engineers can operate the environment after implementation.
You will also work closely with a senior Databricks data-engineering lead who owns the metadata-driven ingestion and data-engineering framework. Your responsibility is to make sure the platform beneath that framework is secure, automated, scalable and production-ready.
What we are looking for
We want someone who has personally done this before.
Strong candidates will have:
8+ years in cloud/platform/data infrastructure engineering or architecture.
Deep hands-on AWS experience.
Strong hands-on Databricks administration and platform engineering experience.
Production experience implementing Databricks on AWS—not simply developing notebooks or pipelines inside an existing workspace.
Advanced Terraform skills and experience managing infrastructure through Git and CI/CD.
Strong AWS IAM and security knowledge.
Strong AWS networking experience: VPCs, routing, PrivateLink, DNS, Transit Gateway and enterprise connectivity.
Practical Unity Catalog experience including catalogs, storage credentials, external locations, grants and workspace/catalog isolation.
Experience with Databricks account/workspace administration, compute policies, serverless and Classic compute.
SSO/SCIM and service-principal implementation experience.
Logging, monitoring, audit and FinOps experience.
Ability to debug infrastructure and platform problems directly rather than hand them to another team.
Strong documentation and communication skills.
Ability to explain architecture trade-offs to security teams, infrastructure engineers and executives.
The comparable Databricks role itself calls for advanced capability in AWS security/identity, VPC/Direct Connect/PrivateLink, HA/DR/observability, Terraform or equivalent IaC, Python/SQL and senior-level architecture discussions.
Particularly valuable experience
Databricks Platform Administrator or related Databricks certifications.
AWS Solutions Architect Professional or equivalent.
Healthcare, financial services or another regulated-data environment.
HIPAA/PHI environments.
Enterprise PrivateLink deployments.
Databricks serverless networking.
Large Unity Catalog implementations.
Terraform module design rather than only consuming existing modules.
Databricks Asset/Declarative Automation Bundles.
CloudWatch, CloudTrail and SIEM integration.
Experience working where another managed-services provider controls portions of the AWS estate.
Experience creating implementation-ready change packages for infrastructure/security teams.
Python and SQL.
AI-assisted engineering using tools such as Claude Code, Codex or equivalent—with human review and infrastructure controls.
Databricks itself lists Databricks and AWS certifications as desirable for comparable platform architecture positions.
What success looks like
In your first major implementation, you should be able to take an approved AWS/Databricks architecture and help turn it into:
Dev and Preprod environments that are stable and reproducible; private connectivity proven; identity, secrets and service principals working; Unity Catalog controls enforced; CI/CD operational; audit events visible; cost controls active; and a client engineering team capable of deploying the reference workload without depending on you.
How we work
We believe infrastructure should be code, not clicks. Terraform, configuration, tests, readiness probes, deployment automation and runbooks live in Git. AI coding tools can accelerate implementation, testing and documentation, but every material infrastructure or security change has a named human reviewer.
We are a small senior team. This role will have substantial autonomy and direct access to company and client leadership.
Location
United States, remote.
Some travel to client locations may be required.
Because some client environments may contain regulated healthcare information, access to specific environments may be subject to client security, location and authorization requirements.
Compensation
Competitive and based on experience and employment structure.

Work arrangement
No