Principal Machine Learning Architect | Enterprise Agentic Search

Overview

Working at Atlassian

Atlassians can choose where they work – whether in an office, from home, or a combination of the two. That way, Atlassians have more control over supporting their family, personal goals, and other priorities. We can hire people in any country where we have a legal entity.

Your future team

Rovo helps people and AI agents find, understand, and act on enterprise knowledge. We build the search and machine-learning capabilities that connect agents with relevant information across Atlassian products and connected applications.

Enterprise search presents a distinctive challenge for agents. Information is scattered across documents, conversations, tickets, and other systems. Questions depend on company-specific terminology, relationships, permissions, and information that changes over time. Agents need to discover the right sources, refine their searches as they learn, and gather enough reliable evidence to complete a task efficiently.

We are hiring a Principal Architect to define how enterprise search powers AI agents. You will own some of the followings : the architecture and technical direction connecting retrieval, agent orchestration, context construction, evaluation, and large language model (LLM) training. Your work will help agents search effectively, gather reliable evidence, and reason over enterprise knowledge across products and applications.

This is a hands-on architect role with impact across multiple teams. You will make foundational design decisions, validate them through prototypes and production evidence, and guide engineers and scientists through implementation and adoption. You will establish a coherent architecture that teams can evolve as models, enterprise data, and customer needs change.

Responsibilities

You will be working on some of the following areas:

Own the architecture for enterprise agentic search. Define how query understanding, source discovery, retrieval, agent orchestration, tool use, and context construction work together. Establish system boundaries, interfaces, and reusable capabilities that support multiple agent experiences.

Design for enterprise complexity. Make architectural decisions for heterogeneous content, permissions, tenant isolation, freshness, domain-specific language, and information needs spanning multiple systems. Balance search quality and agent effectiveness with scalability, latency, cost, and reliability.

Architect evaluation and experimentation. Define the benchmarks, datasets, evaluation environments, and observability needed to assess retrieval quality, agent search behavior, evidence coverage, and task success. Guide teams in building reproducible experiments and calibrated evaluators that inform architecture, model, and launch decisions.

Set the direction for training LLMs for agentic search. Define model capabilities, learning objectives, training data requirements, and the architecture connecting training, evaluation, and serving. Lead technical decisions on supervised fine-tuning, preference optimization, and reinforcement learning to improve search planning, tool use, iterative evidence gathering, and grounded reasoning. Partner with engineers and scientists to train, validate, and deploy these improvements.

Design the continuous improvement loop. Connect production interactions, agent trajectories, and failure analysis to evaluation cases, training data, and model updates. Establish data and feedback interfaces that preserve permissions, privacy, and evaluation integrity.

Validate architecture through implementation. Build prototypes and reference implementations, investigate difficult system failures, and work directly with teams on critical components. Use experiments and production results to resolve trade-offs and guide incremental adoption.

Lead technical direction across teams. Set a multi-quarter roadmap, align search, agent, and ML platform teams on shared architecture, and guide implementation through design reviews and technical mentorship. Take accountability for architectural outcomes and measurable improvements in agent task success.

Compensation

At Atlassian, we strive to design equitable, explainable, and competitive compensation programs. We follow consistent hiring practices and account for each candidate's skills, knowledge, and experience when setting base pay within the range.

Please visit go.atlassian.com/payzones for more information on which locations are included in each of our geographic pay zones. However, please confirm the zone for your specific location with your recruiter.

This role may also be eligible for benefits, bonuses, commissions, and equity.

Pay Ranges

In The United States , we have three geographic pay zones. For this role, our current base pay ranges for new hires in each zone are:

Zone A : $236,700 - $309,025

Zone B : $213,030 - $278,123

Zone C : $196,461 - $256,491

Qualifications

What we are looking for

Substantial hands-on industry experience architecting, delivering, and evolving production RAG or agentic search systems used by real customers, with particular depth in enterprise applications.

A record of owning architecture from ambiguous customer needs through system design, implementation, launch, and evolution, with measurable improvements in search quality, task success, or customer outcomes.

Demonstrated ability to define system boundaries, interfaces, and shared capabilities across teams, and to make sound architectural decisions about scalability, reliability, extensibility, and operational complexity.

Practical experience with enterprise knowledge and workflows, including heterogeneous data sources, access permissions, freshness, domain-specific terminology, and complex information needs spanning multiple systems.

Deep understanding of LLM-based search and agent systems, including retrieval, tool use, multi-step search, context construction, and grounded reasoning, and how these components affect end-to-end performance.

Practical experience adapting and deploying LLMs, with the technical depth to guide training data design, supervised fine-tuning or other post-training methods, evaluation, and serving decisions. Experience improving search or tool-use behavior is especially valuable.

Strong evaluation and experimental skills: you can design useful benchmarks, assess label and judge quality, recognize leakage and confounding factors, and connect offline results to product impact.

Strong software and distributed-systems engineering skills, with the ability to validate designs in code and work deeply with data pipelines, model serving, and production observability.

A record of influencing technical direction across multiple teams, guiding architecture through production adoption, and mentoring senior engineers and scientists.

Sound judgment about trade-offs across model quality, data availability, latency, cost, reliability, privacy, and access control.

Benefits & Perks

Atlassian offers a wide range of perks and benefits designed to support you, your family and to help you engage with your local community. Our offerings include health and wellbeing resources, paid volunteer days, and so much more. To learn more, visit <a href="http://go.atlassian.com/perksandbe