New opportunity

Applied AI Engineer - ANZ

Cognition · Sydney, Australia

About this role

Cognition seeks an Applied AI Engineer to deploy its AI software engineering agent, Devin, with enterprise customers in Australia. The role combines hands-on engineering, customer enablement and measurable adoption, while helping turn individual engagements into repeatable programs. The position is listed as remote, with Sydney named as its location.

Responsibilities

Embed with engineering teams to integrate agentic workflows into engineering, QA, support, data and product work, and drive lasting adoption of Devin.
Help customers install, configure and optimize Devin and associated tools, including DeepWiki and MCP integrations. Lead live workshops and pair-program on production problems to demonstrate effective usage.
Track productivity outcomes, develop ROI stories and make the case for expanding adoption within customer accounts.
Work with leadership to convert field learnings into structured playbooks, best practices, digital learning content and enablement materials that scale beyond individual engagements.

Requirements

A STEM degree or equivalent hands-on experience, plus at least 3 years in software engineering, technical consulting, deployment strategy, forward-deployed engineering, solutions engineering or a similar role. Strong coding proficiency in Python, JavaScript/TypeScript or a comparable language is required.
Demonstrated ability to explain complex technical topics to varied audiences, drive technical adoption and measurable results within engineering organizations, communicate clearly in writing and verbally, learn quickly, and recognize opportunities to grow customer accounts.

Potential advantages include experience leading developer enablement, platform adoption or internal AI modernization with measurable results; deploying LLM or agent-based systems in production; and working in an early-stage startup requiring autonomy and rapid execution. Estimated base salary is $180,000–$225,000 USD per year, with equity offered.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

PythonJavaScriptTypeScriptAI agentsLLMsMCPDeepWikiDevinPair programmingTechnical consultingDeveloper enablementTechnical communicationCustomer engagementWorkflow designProduct adoptionROI measurement

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

Drive lasting adoption in engineering teams
The role embeds with enterprise teams and owns outcomes beyond onboarding. It involves identifying high-leverage agentic workflows across engineering, QA, support, data and product, then integrating them into how teams work.
Deploy Devin through hands-on enablement
The role guides installation, configuration and optimization of Devin, DeepWiki and MCP integrations. Live workshops and pair programming on production problems are central to demonstrating useful workflows.
Measure impact and support account expansion
Cognition calls for tracking productivity metrics, developing ROI stories and making the business case for broader adoption within customer accounts.
Turn field learning into scalable programs
As one of the first Applied AI Engineers, the hire will work with leadership to turn customer engagements into playbooks, best practices, digital learning content and other enablement materials.

Likely challenges to think through

Balance customer-specific work with repeatability
Preparation hypothesis: workflows tailored to one team's practices may not transfer neatly to others, even as the role must scale field learning. Prepare to explain how you would document reusable patterns while preserving room for customer-specific decisions.
Demonstrate value on live production work
Preparation hypothesis: pair programming on real problems creates a trade-off between showing quick progress and making sound engineering decisions. Prepare a structured approach to choosing a suitable task, checking the agent's output and explaining decisions to the team.
Make productivity gains credible
Preparation hypothesis: adoption and productivity can be difficult to attribute to one tool, yet this role must build ROI stories that support expansion. Prepare to discuss baseline measures, observable usage and the limits of any claimed impact.

Practice your answers

Technical practice

How would you identify the highest-leverage agentic workflow for an engineering team?

What to coverThe role involves finding where AI creates leverage across engineering, QA, support, data and product work. Suggested answer: compare candidate workflows by frequency, effort, risk and measurable outcomes before choosing a starting point.

How would you design an agentic workflow that fits an existing delivery process?

What to coverThe posting calls for architecting and implementing workflows within how teams build and ship. Suggested answer: explain the human and agent responsibilities, review points and how you would test whether the workflow improves the existing process.

How would you guide a customer through deploying Devin and its associated tools?

What to coverThe role includes installing, configuring and optimizing Devin, DeepWiki and MCP integrations. Suggested answer: describe how you would establish the customer's intended use, verify the setup and refine it based on observed usage, without assuming undocumented product features.

How would you troubleshoot an agentic workflow that works in a workshop but not in routine use?

What to coverThe posting distinguishes lasting adoption from onboarding and calls for optimizing deployments. Suggested answer: isolate whether the difficulty lies in configuration, workflow fit or user understanding, then test a change against real team work.

How would you pair-program on a live production problem while demonstrating useful AI usage?

What to coverPair-programming on production problems is an explicit duty, alongside a requirement for strong coding proficiency in Python, JavaScript/TypeScript or a similar language. Suggested answer: walk through scoping the problem, inspecting code, validating a proposed change and keeping the engineer engaged in decisions.

How would you evaluate an LLM or agent-based integration before broader production use?

What to coverProduction experience with LLM or agent-based systems is listed as a potential advantage, not a requirement. Suggested answer: discuss task-specific checks, failure handling and the trade-off between usefulness and the effort needed to review outputs.

How would you measure whether a Devin deployment improves productivity?

What to coverThe role explicitly calls for tracking productivity metrics and quantifying impact. Suggested answer: choose measures tied to the team's work, establish a before-and-after comparison and account for changes that might have occurred without the deployment.

How would you assess whether results justify expanding adoption within an account?

What to coverThe posting asks the engineer to surface ROI stories and make a business case for a larger footprint. Suggested answer: connect observed outcomes to costs and stakeholder priorities, while distinguishing measured results from projections for other teams.

How would you turn a successful customer workflow into a reusable playbook?

What to coverThe role includes converting field learnings into best practices, digital learning content and enablement materials. Suggested answer: separate repeatable steps from customer-specific choices, document how to evaluate results and identify where a new team would need to adapt the approach.

Behavioural practice

Tell me about a time you helped an engineering team adopt a new technical workflow.

What to coverThe posting emphasizes lasting adoption and measurable impact rather than onboarding alone. Suggested answer: give the situation, the actions you took to address adoption barriers and the result, including a real measure if available.

Tell me about a time you explained a complex technical topic to different audiences.

What to coverClear communication with diverse audiences and live workshops are part of the role. Suggested answer: describe the audiences and their starting points, how you changed your explanation and what they were able to do afterward.

Tell me about a time you solved a production problem collaboratively.

What to coverThe role calls for pair-programming on live production problems. Suggested answer: describe the problem, how you worked with the other engineer on decisions and validation, and the outcome without overstating your individual contribution.

Tell me about a time you had to learn quickly and act with limited direction.

What to coverThe posting values fast adaptation and identifies early-stage autonomy as a potential advantage. Suggested answer: explain the unfamiliar situation, how you identified what to learn first, the actions you took and the result.

Tell me about a time you made a customer or team success repeatable.

What to coverThis role involves turning individual engagements into playbooks and enablement materials that scale. Suggested answer: give the original situation, what you documented or changed for other teams, and evidence of whether the approach transferred.

Sharpen your knowledge

Topics to prepare

Agentic workflow design
The role calls for identifying high-leverage uses of Devin across engineering, QA, support, data and product. As preparation, map one team's workflow, choose a task for an agent, and explain where human review and outcome measurement would fit.
Adoption beyond onboarding
The posting emphasizes lasting adoption and measurable impact inside engineering organizations. As preparation, outline how you would distinguish initial usage from sustained use, and what you would change if a team stopped using Devin.
Productivity measurement and account growth
The role tracks productivity metrics, develops ROI stories and makes a case for expanding adoption. As preparation, sketch a before-and-after measurement plan and explain how you would avoid attributing every improvement to Devin.
Scaling customer learnings
Cognition wants field learnings turned into playbooks, best practices and digital learning content. As preparation, identify which lessons from a hypothetical deployment would generalize to other teams and which would remain customer-specific.

Skills to revise

Coding and live pair programming
Strong proficiency in Python, JavaScript/TypeScript or a similar language is required, and the role pairs on live production problems. Practice explaining your choices aloud while implementing and testing a small change in a language you know well.
Devin deployment walkthrough
The role guides customers through installing, configuring and optimizing Devin and associated tools, including DeepWiki and MCP integrations. If you have access to the products, rehearse a configuration walkthrough; otherwise, draft the questions you would ask and the checks you would perform, without assuming specific setup steps.
Interactive technical teaching
The posting calls for live workshops, clear communication to varied audiences and developer enablement. Prepare a short exercise that demonstrates an agentic workflow, then practice explaining its value and limitations to both engineers and engineering leaders.
Impact reporting
The role requires tracking productivity and communicating measurable results. Using hypothetical data, create a concise before-and-after summary that states the metric, baseline, observed change and uncertainty, then turn it into a defensible ROI story.
Reusable enablement writing
The role creates shared playbooks and enablement materials from customer engagements. Draft a one-page guide for repeating a hypothetical workshop, including prerequisites, steps, success measures and lessons to revisit.

Your preparation checklist

Select a technical project walkthrough
The role requires strong coding proficiency. Prepare a concise walkthrough of a Python, JavaScript/TypeScript or comparable project, showing your implementation decisions and how you resolved a difficult issue.
Sketch an agentic workflow
As a practice exercise, choose an engineering, QA, support, data or product task and diagram how an agent could help, where human review belongs and how you would judge whether the workflow helps.
Prepare a deployment discussion
The role involves installing, configuring and optimizing Devin, DeepWiki and MCP integrations. Outline how you would learn a customer's environment, plan an integration and troubleshoot it; distinguish direct experience from your proposed approach.
Plan a hands-on enablement session
The posting calls for live workshops and pair programming on production problems. Draft a short session agenda that includes a practical task, audience-appropriate explanations and a way to check what the team can use afterward.
Assemble evidence of adoption and impact
Prepare an example, if available, with a before-and-after measure, your contribution and how you encouraged continued use. The role emphasizes lasting adoption, productivity metrics, ROI stories and account expansion; avoid claiming outcomes you cannot substantiate.
Draft a reusable field playbook
As a preparation exercise, turn one customer-engagement example into a one-page outline covering discovery, workflow selection, enablement, measurement and lessons that another team could reuse.
Bring focused questions for Cognition
Ask how Applied AI Engineers prioritize workflows, measure durable adoption, turn field learning into shared materials and coordinate with leadership on expanding successful engagements.

Availability last checked: 4 Nov 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in Australia. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…