About this role
Senior Staff Software Engineer, Agentic Applications
Enterprise Platform Group · Commercialization (Quote-to-Cash) · Mountain View, CA
As a Senior Staff Software Engineer on the Agentic Applications team, you will be the hands-on technical lead for how Databricks quotes, orders, bills and recognizes revenue. Your core function is to build bespoke first-party, agents-first applications built on the Databricks Data Intelligence Platform. You will own the technical architecture end to end, from quote to cash. You will design agents that work reliably inside revenue-critical workflows, set the engineering standards the team builds on, and make sure every agent-assisted decision can be observed, audited and trusted by Finance, Deal Desk and Sales.
This is a hands-on technical leadership role, not a people-management role. You will write code, review designs, and be the most senior technical voice in a domain that touches every dollar Databricks books.
The impact you'll have
• Own the agentic architecture for Quote-to-Cash. Define and lead the target technical architecture across Configure-Price-Quote, deal management, ordering, subscriptions, investments, billing and revenue recognition. Drive the move from vendor-configured workflows to first-party, AI-native applications built on Databricks Apps, Lakebase, Unity Catalog and Mosaic AI.
• Build production agents that sellers and Deal Desk trust. Design and ship LLM-powered agents for deal guidance, quote generation and summarization, multi-option proposals, approval insights, and pricing and discount recommendations. Work directly with sales, Deal Desk and Finance to define what "good" looks like for each workflow.
• Make agent reliability a core system property. Own the eval and quality system for this domain: per-workflow quality criteria, offline eval suites, LLM-as-judge and human-labeled benchmarks, behavioral regression testing in CI/CD, online tracing, and feedback loops that improve agents based on what users approve, edit and reject.
• Design human-in-the-loop controls for revenue-critical decisions. Build the separation between what an agent proposes and what gets committed, plus diff-based review, approval routing, audit logging, and policy guardrails. Together these decide when an agent can act on its own and when a person has to approve, especially for pricing, discounting, contract terms and order creation.
• Treat systems of record as agent tools. Architect tool-calling and MCP-style interfaces that give agents governed, run-time access to pricing, product catalog, order, contract, subscription and billing data. Follow single-source-of-truth principles across CRM, ERP and billing systems.
• Protect financial integrity at scale. Make sure agent-assisted quoting and ordering produce accurate, compliance-grade outputs downstream: order validation, billing, invoicing and revenue recognition, where errors directly affect financial reporting.
• Lead the migration path. Act as th