Keirsten Keys is a second-career attorney and legal-tech engineer specializing in the design, execution and validation of AI-assisted workflows and work product relied upon in high-stakes litigation. Before entering law, she spent more than a decade running businesses of her own, including co-founding a car dealership, managing an auto loan brokerage and operating a digital marketing agency serving consumer and B2B clients across various industries. Keirsten’s background shapes her approach to legal practice; she brings an operator’s eye to high stakes matters, developing legal strategies around what scales, what breaks and what holds up under real-world circumstances.

Keirsten specializes in prompt engineering and development of custom legal AI tools calibrated to the specific facts, issues and constraints of each matter. She has designed and deployed AI-assisted eDiscovery, document review, triage and analysis solutions engineered for reliability and defensibility from the outset. She is skilled at translating fragmented, high-volume data into organized, actionable work product, and brings her own methodologies for evaluating and red-teaming AI-generated content for veracity, consistency and domain-specific accuracy across legal, business and technical contexts.

Her broader competencies include iterative prompt refinement, process optimization, RLHF-aligned quality assessment, high-volume dataset analysis and document synthesis. She also engineers made-to-order AI drafting tools calibrated to produce consistently structured outputs with contents intelligently adjusted to the unique facts of each run. Keirsten custom-designs and optimizes AI-based solutions that reduce time spent on clerical, due diligence and other repetitive tasks, freeing attorneys to focus their attention on tasks where high-level strategic thinking, skill and experience matter most and deliver the highest value to clients.

As a complement to her AI skills, Keirsten brings deep technical fluency across the wider legal tech stack. She combines her literacy in Python, HTML and other coding languages with advanced software proficiency to provide offerings that leverage features such as task automation, Excel formulas and cross-referencing and eDiscovery platform and CRM data mining to develop truly unique, one-of-a-kind legal solutions. The end result is workflows, pipelines, trackers and other systems that effectively streamline matters marked by complex multi-variate issues and provide litigation teams with a single, centralized source of truth to rely upon.

At Shook, some of Keirsten’s AI-assisted contributions include:

  • Designing and executing AI-assisted document review workflows for large-scale litigation discovery, processing thousands of documents across multiple matters for a Fortune 100 technology client.
  • Leveraging AI alongside custom Python code to execute complex eDiscovery and review tasks, including data analysis, production deficiency checks, scoring and ranking and triage.
  • Authoring custom prompt instructions and issue-framing prompts that guide AI tools toward legally relevant distinctions, replacing generic summarization with targeted legal analysis.
  • Testing AI outputs against independent legal judgment, identifying precision gaps, adjusting review logic, and re-executing to produce repeatable, defensible, and auditable results.
  • Building structured frameworks that pair AI-generated outputs with manual verification and tracking documentation to create transparent audit trails demonstrating how analytical conclusions were reached.
  • Synthesizing AI-assisted review outputs into attorney-ready deliverables that translate raw data and complex technical analyses into actionable strategic insights for clients, in-house counsel and litigation teams.
  • Developing systematic tracking infrastructure for multidistrict litigation, consolidating fragmented datasets into a single auditable source of truth relied upon across the litigation team.
  • Creating multiple AI-automated drafting systems that deliver consistent, fact-adaptive work product at scale, compressing drafting cycles and freeing senior attorneys to focus on judgment and strategy.

What unifies Keirsten’s work is the discipline of treating AI as defensible infrastructure rather than novelty: she builds, tests, refines and audits AI-assisted solutions meticulously before they touch a client matter or circulate to any legal team. She brings an entrepreneur’s instinct for what’s worth building, a litigator’s instinct for what holds up under scrutiny and an engineer’s discipline for proving the difference. Amid AI’s rapid transition from experimental novelty to commonplace device, Keirsten is positioned at the cutting-edge, helping shape the future of AI-assisted legal practice through developing the highly-customizable, trusted legal tech solutions of the future.