About this role.
Zuora is hiring a Senior Software Engineer to architect and scale an agentic deal-desk platform in a hybrid Foster City, California role. The engineer will build Java microservices and the AI agent harness, including execution loops, tool dispatch, state management, prompt assembly, validation, and retries. The role requires deep expertise in Java, distributed systems, cloud infrastructure, secure runtime design, and LLM-agent orchestration. It also carries technical leadership expectations, including setting engineering standards, advancing evaluation-driven development, and mentoring senior engineers.
Role DNA
A quick view of the complexity, pace, ownership and collaboration implied by the job description.
Job Complexity
5/5Pace & Pressure
5/5Autonomy Level
5/5Communication Load
4/5Salary analysis
Estimated compensation compared with the broader US market for similar roles.
Core skills
Skills and capabilities most closely associated with this opportunity.
Sample interview questions
I would separate model output from execution through strict schema validation, authorization checks, deterministic policy rules, and a bounded retry strategy. Each action would be logged with correlation IDs and state transitions, while unsafe or invalid requests would be rejected or routed to a safe fallback rather than executed.
I would keep a compact, structured session state for current deal facts and decisions, store auditable events separately, and use retrieval for durable historical knowledge. Context compaction would preserve constraints, approvals, and unresolved actions while vector retrieval would supply only relevant policy or account information to each step.
I would prioritize idempotent consumers, durable event contracts, ordering strategy, retry and dead-letter handling, and observability across asynchronous workflows. I would use synchronous APIs only where immediate user feedback is required, while Kafka or similar messaging would support resilient state transitions, audits, and integrations.
Tool permissions should be enforced by the runtime, not trusted to the model prompt. I would use allowlisted tools, least-privilege credentials, typed input schemas, policy checks against trusted system data, sandboxed execution, and complete audit trails for every attempted action.
I would define representative golden datasets and measurable criteria such as tool-selection accuracy, policy compliance, latency, cost, and task completion rate. In production, I would trace every run, monitor failure patterns and drift, sample outcomes for human review, and use controlled releases to validate changes before broad rollout.
About Zuora
At Zuora, we help businesses grow smarter and adapt faster. Our platform powers modern business models — from subscriptions and usage-based pricing to AI-driven and outcome-based offerings — helping companies launch new products, automate complex billing, and unlock predictable, recurring revenue.
We’ve led the Subscription Economy for more than a decade. Now we’re evolving again by building the definitive platform for quote to cash and helping companies monetize their products and services with an adaptable, AI-ready foundation.
Role Overview
Zuora is seeking a Senior Software Engineer to architect and scale our Agentic Deal Desk product. In this high-impact IC role, you will build resilient, high-throughput Java microservices alongside the underlying AI agent harness—the runtime scaffolding, execution loop, tool dispatch system, and state management engine that converts non-deterministic LLMs into reliable, enterprise-grade deal-desk agents.
Location & Work Arrangements
This role is based in our Foster City, CA office and follows a hybrid work model. Team members are expected to work onsite three days per week to foster collaboration, innovation, and team connectivity.
This is not a fully remote position.
Key Responsibilities
- Agent Harness Engineering: Build and own the core agent harness layer, including the execution loop (ReAct/Plan-Act), state persistence, dynamic prompt assembly, output schema validation, and retry mechanisms.
- Tool Dispatch & Sandboxing: Design safe tool-execution runtimes and schema-validated function dispatchers that allow agents to execute complex deal operations, policy checks, and pricing adjustments securely.
- Context & Memory Management: Engineer context compaction, short-term session scratchpads, and long-term memory retrieval pipelines (RAG/vector search) to eliminate context degradation during multi-turn deal negotiations.
- Guardrails & Feedback Loops: Build computational and inferential feedback sensors (evaluators, deterministic policy linters, and verification checks) to enforce strict margin guardrails, discount limits, and compliance rules.
- Backend Architecture: Design low-latency, event-driven Core Java microservices to back the harness, driving state transitions, audit logs, and integration with Zuora Billing and CRM platforms.
- Technical Leadership: Set technical standards for agent harness design, drive evaluation/eval-driven development practices, and mentor senior engineers across squads.
Required Qualifications
- Core Java Mastery: 8+ years of hands-on experience in modern Java (17+), JVM internals, concurrency models, Spring Boot, and cloud-native distributed architecture.
- Agent Harness & Runtime Systems: Deep experience building the infrastructure surrounding LLMs—agent loops, state-machine tracking, tool calling/dispatch systems, context window management, and structured output parsing.
- Agentic Frameworks & AI Tooling: Practical experience with LangGraph, CrewAI, AutoGen, LlamaIndex, or custom in-house agent orchestration runtimes and vector databases (e.g., Pgvector, Pinecone).
- Distributed Infrastructure: Heavy background in event-driven architecture (Kafka/RabbitMQ), high-performance caching (Redis), relational databases (PostgreSQL/MySQL), and REST/gRPC API design.
- Cloud & Runtime Security: Experience with AWS, Docker/container isolation, Kubernetes, and securing agent execution pathways against improper tool invocations or prompt injection.
Preferred Experience
- AWS Agentic Enablement: Production experience leveraging Amazon Bedrock Agents (Action Groups, Knowledge Bases, Multi-Agent Collaboration) and Guardrails for Amazon Bedrock to enforce policy boundaries and tool permissions.
- Serverless Agent Scaffolding: Experience utilizing AWS Lambda, Step Functions, DynamoDB, and ECS/Fargate for isolated tool-execution environments, persistent session state checkpointing, and agent workflow orchestration.
- AWS AI Observability & Evals: Familiarity with continuous agent monitoring, trace/span evaluation frameworks (e.g., Amazon Bedrock AgentCore Evaluations, CloudWatch), and LLM-as-a-judge pipelines for non-deterministic applications.
- Domain Expertise: Experience in Enterprise Deal Desk operations, revenue operations, quote-to-cash workflows, dynamic pricing engines, or high-volume SaaS approval chains.
Benefits
Zuora offers a comprehensive total rewards package designed to support ZEOs’ wellbeing, growth, and flexibility. While specific offerings may vary by country, we typically provide:
- Competitive compensation, variable bonus and performance-based reward opportunities, and retirement programs
- Medical Insurance
- Generous, flexible time off, plus paid holidays, wellness days, and a company-wide year-end break
- Paid parental leave (including fully paid leave for eligible ZEOs, subject to local policy)
- Learning & development stipend to support ongoing growth
- Opportunities to volunteer and give back, including charitable donation matching where available
- Mental wellbeing resources and support
*Benefits may vary by location; details will be shared during the interview process
#ZEOLife at Zuora
ZEOs (our employees) are empowered to take ownership, challenge the status quo, and make a real impact. We:
- Collaborate deeply across teams and regions
- Learn constantly and iterate often
- Build an inclusive, high-performance culture where people feel inspired, connected, and valued
Our Commitment to an Inclusive Workplace
Think, be and do you.
At Zuora, different perspectives, experiences, and contributions matter — everyone counts.
Zuora is proud to be an Equal Opportunity Employer committed to creating an inclusive environment for all. We do not discriminate on the basis of, and consider individuals seeking employment with Zuora without regard to, race, religion, color, national origin, sex (including pregnancy, childbirth, reproductive health decisions, or related medical conditions), sexual orientation, gender identity, gender expression, age, status as a protected veteran, status as an individual with a disability, genetic information, political views or activity, or other applicable legally protected characteristics.
We encourage candidates from all backgrounds to apply. Applicants in need of special assistance or accommodation during the interview process or in accessing our website may contact us by sending an email to assistance@zuora.com (or local equivalent, where applicable).
The base pay details represent the annualized salary range for the posted position. While we share a comprehensive range, a candidate’s final base salary will fall within these guidelines and will be determined based on multiple factors including but not limited to: qualifications of the candidate, job related knowledge, prior related experience, specific and unique skills, the location of the role, internal equity and internal budget. Base pay is only one element of Zuora’s Total Rewards Package. Additional elements may also include bonus/variable compensation, equity and comprehensive benefits, more details found here.
Base Pay Details
$164,800—$226,600 USD
Upload your resume now
- To unlock remote work opportunities and be discovered by global employers.
This job listing has been manually reviewed by the Jobicy Trust & Safety Team for compliance with our posting guidelines, including verification of the company's legitimacy, accuracy of job details, clarity of remote work policy, and absence of misleading or fraudulent content.
