New opportunity

AI Innovation Advisor

Kirkland & Ellis · New York, United States

About this role

The AI Innovation Advisor works within Kirkland & Ellis’s Innovation Program AI Advisory Services team, embedding with legal practice groups to turn attorney tasks into AI-assisted workflows. The role partners with engineers, data scientists, project managers and practice teams, and may support AI-enhanced services delivered directly to clients. It reports through matrixed management to the Associate Director of Innovation Engineering.

Responsibilities

Advise litigation and transactional teams on AI integration; document existing workflows, define requirements, identify opportunities to reuse existing solutions, and coordinate stakeholders and timelines.
Own end-to-end legal AI workflow development: break down legal tasks, design and test prompts, benchmark outputs against firm standards, and refine workflows with attorney feedback. Keep decision logs covering requirements, prompt logic, validation criteria, failure modes and platform constraints so workflows can be reconstructed on future infrastructure.
Partner with engineers as workflows mature. Contribute templates, decision trees and implementation guidance to a scalable AI workflow playbook.
Explain generative AI capabilities, limitations and firm governance requirements; develop demonstrations, user guides and training with training teams. Report project status, adoption metrics and outcomes, and stay current on AI developments and regulatory changes.

Qualifications

A J.D. from an accredited law school is strongly preferred. Candidates without a law degree who have five or more years leading innovation initiatives, technology adoption or AI implementation at a large law firm or legal technology company will also receive strong consideration.
Experience building and iterating prompts and workflows on legal AI platforms such as Harvey, Legora, CoCounsel or Lexis+ AI, with accountability for output quality. The role requires an understanding of legal technology, generative AI and large language models, plus strong writing, project management, organizational, collaboration and stakeholder-advisory skills. Candidates must be comfortable with the quality, pace and confidentiality expectations of billable client work.
Preferred qualifications include three years as an attorney at an AMLAW 50 firm or comparable legal environment, experience training legal teams, and knowledge of data security, information governance and AI compliance.

The posting is tagged hybrid. The New York base salary range is $153,000–$211,000; listed benefits include healthcare, paid time off, retirement, and learning and development opportunities.

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

Skills for this role

Generative AILarge Language ModelsPrompt EngineeringLegal AI PlatformsHarveyLegoraCoCounselLexis+ AIAI Workflow DesignAI TestingOutput BenchmarkingResponsible AILegal TechnologyProcess ImprovementProject ManagementStakeholder ManagementAttorney TrainingData SecurityInformation GovernanceCompliance

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

Scope AI opportunities with practice groups
The role embeds with litigation and transactional teams to document current workflows, translate attorney tasks into requirements, and identify existing builds or known failure points. Prepare to explain how you would distinguish a useful AI workflow from a task better handled another way.
Own workflow quality from design through testing
The posting calls for decomposing legal tasks, building and iterating prompts, benchmarking outputs against firm standards, and refining workflows with attorneys. Prepare a concrete approach to validation, failure analysis, and deciding when an output is ready for use.
Make workflows reproducible and scalable
The advisor maintains decision logs covering requirements, prompt logic, validation criteria, failure modes, and platform constraints, then partners with engineers and contributes to a shared playbook. Prepare to describe documentation that would let another team reconstruct a workflow on future infrastructure.
Guide responsible adoption
The role explains AI capabilities, limitations, and firm governance requirements; works with training teams on demonstrations and guides; and reports adoption metrics and outcomes. Prepare to connect training and feedback to changes in actual attorney workflows.
Deliver in a client-service setting
The advisor may support AI-assisted services delivered directly to clients and must meet the pace, quality, and confidentiality expectations of billable work. Prepare to discuss how you would coordinate attorneys, technical partners, and timelines without weakening review standards.

Likely challenges to think through

Speed versus defensible output quality
Likely preparation hypothesis, not a reported firm problem: rapid experimentation may conflict with the validation needed for legal or client-facing work. Prepare a staged testing approach with attorney review, explicit acceptance criteria, and a way to record failures before wider use.
Local fit versus reuse
Likely preparation hypothesis: practice groups may need tailored workflows, while the role must also reuse existing builds and contribute to a scalable playbook. Prepare to explain which task requirements warrant customization and which components could remain shared.
Adoption versus governance
Likely preparation hypothesis: making AI useful to busy attorneys may require simple guidance, while its limitations, confidentiality expectations, and firm governance still need to be understood. Prepare a training approach that pairs practical demonstrations with clear boundaries and a feedback mechanism.

Practice your answers

Technical practice

How would you turn an attorney’s existing task into a scoped AI workflow?

What to coverThe role documents litigation and transactional workflows and translates attorney needs into requirements. Suggested preparation: explain how you would map the current task, identify quality and confidentiality constraints, define a useful outcome, and check whether an existing build can be reused.

How would you break down a complex legal task before applying generative AI?

What to coverThe posting calls for decomposing legal tasks during end-to-end workflow development. Suggested preparation: distinguish steps suited to AI assistance from steps requiring attorney judgment, and explain how you would define inputs, outputs and review points.

How would you design and iterate prompts for a legal AI workflow?

What to coverHands-on prompt and workflow iteration on legal AI platforms is stated in the posting. Suggested preparation: describe a testable prompt design, how you would compare revisions, and how attorney feedback would guide changes without sacrificing output quality.

How would you benchmark AI outputs against firm standards?

What to coverThe role designs testing protocols and aligns outputs with firm standards. Suggested preparation: explain how you would set validation criteria, review a representative range of outputs with attorneys, and decide whether a workflow is ready for use.

How would you investigate a recurring workflow failure?

What to coverThe posting requires documenting failure modes and partnering with engineers on failure analysis. Suggested preparation: explain how you would reproduce the issue, isolate whether it stems from task design, prompting or platform constraints, and retest a proposed fix.

What would you record so a workflow could be reconstructed on future infrastructure?

What to coverStructured decision logs must capture requirements, prompt logic, validation criteria, failure modes and platform constraints. Suggested preparation: explain how those records would make design decisions and test results understandable to someone rebuilding the workflow.

When should a maturing workflow be handed off to engineers?

What to coverThe advisor owns workflow development and partners with engineers as workflows mature. Suggested preparation: discuss what evidence and documentation would support implementation, and how you would communicate attorney requirements and unresolved limitations.

How would you evaluate whether a legal AI platform fits a proposed task?

What to coverThe role makes tool recommendations and iterates across platforms; the posting names Harvey, Legora, CoCounsel and Lexis+ AI as examples of relevant experience. Suggested preparation: compare platforms against the task’s output quality, constraints and governance needs rather than assuming one tool fits every workflow.

How would you explain responsible AI use for client-facing legal work?

What to coverThe posting emphasizes generative AI limitations, firm governance and the confidentiality and quality expectations of client work. Suggested preparation: explain how you would communicate those limits, establish appropriate attorney review and verify that a proposed use complies with applicable firm requirements.

How would you tell whether a deployed workflow is being adopted effectively?

What to coverThe role develops demonstrations and guides, gathers feedback, and reports adoption metrics and outcomes. Suggested preparation: distinguish usage from useful outcomes and explain how feedback and reported results would inform training or workflow changes.

Behavioural practice

Tell me about a time you earned a senior stakeholder’s trust while advising on a new approach.

What to coverThe role calls for credibility with attorneys and senior stakeholders in ambiguous settings. Suggested preparation: use a situation, your actions and the result to show how you understood their work, addressed concerns and influenced a decision.

Describe a time you coordinated several stakeholders and competing deadlines.

What to coverThe posting calls for managing concurrent initiatives, timelines and stakeholders in a matrixed environment. Suggested preparation: give a situation, the actions you took to clarify priorities and responsibilities, and the resulting delivery or trade-off.

Tell me about a time critical user feedback changed a solution you had built.

What to coverAttorney-driven refinement is central to the advertised workflow-development work. Suggested preparation: describe the original situation, how you evaluated and acted on the feedback, and what changed in the result.

Describe a time you maintained quality under client-service pressure.

What to coverThe role requires sound judgment amid the pace, quality and confidentiality expectations of billable client work. Suggested preparation: use a real situation, the steps you took to protect standards while meeting a deadline, and the outcome; avoid disclosing confidential details.

Tell me about a time you helped a professional team adopt a new tool or process.

What to coverThe posting includes attorney training, demonstrations, user guides and feedback mechanisms. Suggested preparation: explain the situation, how you tailored your communication and training, and what observable result showed whether adoption improved.

Sharpen your knowledge

Topics to prepare

Litigation and transactional workflow fit
The posting calls for embedding with both practice groups and identifying where existing AI solutions can be reused. To revise, compare a hypothetical litigation task and transactional task: identify the attorney decision points, possible AI assistance, and points requiring human review.
Generative AI limitations in legal work
The role requires explaining where large language models succeed and fail while meeting client-work quality standards. To revise, explain how an AI output could be incomplete, unsupported, or misleading, and what an attorney would need to verify before using it.
Confidentiality and AI governance
The posting requires attention to client confidentiality and firm governance; data security, information governance, and AI compliance knowledge are preferred. To revise, outline questions to resolve before using client material in an AI workflow, without assuming Kirkland’s specific policies.
Reusable and portable workflow design
The role calls for decision logs detailed enough to reconstruct workflows on future infrastructure and for a scalable playbook. To revise, identify which task requirements, prompt logic, validation criteria, failure modes, and platform constraints must survive a platform change.

Skills to revise

Scope an attorney request
The posting requires documenting current-state workflows, defining requirements, and coordinating stakeholders. Practice a mock discovery conversation, then produce a one-page scope with inputs, desired output, reviewers, known failure points, reuse opportunities, and open decisions.
Iterate a legal AI prompt
Hands-on prompt and workflow development on Harvey, Legora, CoCounsel, Lexis+ AI, or comparable platforms is required. Using a platform if available, or a written mock-up otherwise, draft and revise a prompt for a hypothetical legal task; record what changed and why.
Benchmark outputs and incorporate feedback
The role owns testing protocols, output benchmarking against firm standards, and attorney-driven refinement. For a fictional task, create a review checklist covering accuracy, support, completeness, and usable format; score sample outputs and describe the next revision. Do not presume knowledge of Kirkland’s internal standards.
Document a reproducible handoff
The posting calls for structured decision logs and collaboration with engineers as workflows mature. Draft a short handoff capturing requirements, prompt logic, validation results, failure cases, platform constraints, and unresolved engineering questions.
Teach and report on adoption
The role includes demonstrations, user guides, feedback mechanisms, and reporting adoption metrics and outcomes. As preparation, sketch a five-minute attorney demonstration and a brief status update distinguishing usage, user feedback, and verified workflow outcomes; these are practice exercises, not stated firm templates.

Your preparation checklist

Map your experience to the role
Prepare a one-page evidence sheet covering legal AI platform work, prompt and workflow iteration, output quality, stakeholder advising, and project management. Note the posting’s J.D. preference and alternative experience pathway without treating either as an eligibility verdict.
Build a legal workflow walkthrough
Prepare a sanitized example showing an attorney task, the current workflow, scoped AI requirements, opportunities to reuse an existing solution, and the handoff to engineers. Use litigation or transactional work if relevant to your experience.
Rehearse a workflow design exercise
Practice decomposing a hypothetical legal task into inputs, AI-assisted steps, attorney review points, and expected outputs. Explain how you would choose where AI adds value and where it should not be used.
Show how you test and document quality
Prepare a sample test plan and decision log covering prompt logic, output benchmarks, validation criteria, failure modes, platform constraints, and changes made after attorney feedback. Use fictional or approved material, not confidential client work.
Prepare a responsible-adoption explanation
Draft a brief attorney-facing demonstration or user-guide outline that explains generative AI capabilities and limitations, confidentiality concerns, and how users would follow applicable firm governance requirements.
Outline delivery and measurement
Prepare an example project-status update showing stakeholders, decisions, timelines, risks, and proposed adoption and outcome measures. Explain how you would coordinate practice teams, engineers, and training partners across concurrent initiatives.
Bring targeted employer questions
Ask how practice groups prioritize workflow opportunities, what firm standards govern output validation, when a workflow moves to engineering, and how the team measures adoption and client-facing outcomes.

Availability last checked: Nov 4, 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 United States. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…