- End-to-end solution architecture across ingestion, warehousing, semantic layers, and conversational delivery, with disciplined architecture decision records and trade-off analysis.

- Azure platform depth: Azure SQL tiering and DTU/vCore economics, private networking and single-tenant security patterns, Azure OpenAI (or equivalent) deployment architecture, and identity and access design.

- Capacity and performance architecture: concurrency modelling, bottleneck diagnosis from measured evidence, and remediation design across tuning, caching, pre-aggregation, read scaling, and tier changes.

- Governed natural-language analytics design: intent taxonomies, parameterized query function contracts, versioning and access control, and the architectural case for constrained function selection.

- Answer fidelity architecture: designing where truth is established and how it is protected through delivery, including containing or eliminating channel-side generative rewriting.

Preferred Candidate Profile:

- At least 10+ years in engineering roles, including 4+ years as a solution or data architect accountable for end-to-end designs in production.

- Proven architecture ownership of a data-and-AI platform on Azure, including warehouse design, capacity planning, and an LLM component deployed in the customer's own tenant.

- Experience architecting integrations across a Salesforce-ecosystem estate plus operational sources such as contact centers or survey systems.

- Experience presenting architecture options and trade-offs to executive stakeholders and defending recommendations when challenged.