Job Description Alignment

The four mandates of the role, mapped to proven work.

Why
I love language, as a professional skill, but even more as a human thing.
Clarity, precision, the work of sharing exactly what I mean with exactly the right words.
When people talk about working with generative AI, about prompting, about outputs, what they're really talking about is communication.
Semantics.
The same discipline we've always needed to do well with each other.
The technology is new.
The underlying work isn't.
That's why I enjoy working with legal teams.
They think in language the way I think in language.
A poorly chosen word has consequences.
Ambiguity has consequences.
And it's why governance is the foundation of everything I build with AI.
I want to know the output is right, traceable, explainable and auditable.
That's a language instinct. I want to know if the words are true.
What
I find the highest-value AI use cases, run POCs to prove them, track what's working, and teach people to use the tools.
Most people do one or two of those things. I do all four.
I bridge the group that has the problem, the team that builds the solution, and the data that proves the value.
85% of firms don't track AI ROI — I've built the systems that put a firm in the 15% that does.
I understand how law firms make money — and where AI creates tension with that model.
01 Builder
02 Tracker
03 Teacher
04 Protector
MANDATE 01

The Builder

AI Strategy & Solution Design
Strategy and execution. Find the opportunity. Translate it into requirements. Prove it with a POC. Deploy with the technology team.
What the JD asks
Partner with practice groups to spot AI opportunities in client workstreams
Translate legal and business needs into AI solution requirements; facilitate proof-of-concept development
Design and deploy scalable AI tools with the technology, data, and e-discovery teams
What I have done
Discovery-first delivery on live legal work
Ran listening sessions and workflow mapping across Legal, Procurement, Finance, and People Operations. Shipped agents against real infrastructure constraints: a separation agreement generator across US, EMEA, and Australia, and a legal metrics agent on contract execution and outside-counsel spend data.
Fortune 500 SaaS · 4 departments
POC-to-production on Glean and n8n
5 agents deployed in 6 months across departments that had never used AI. Each started as a POC, graduated after stakeholder review. ~190 hrs/year saved.
Enterprise engagement
AESOP OS: the POC pipeline, productized
990 tests. 6 evaluation modes. 30+ workflows. Turns discovery into structured requirements and PRDs. POC pipeline tracks every initiative from Discovery → Build → Review → Graduated.
AESOP · Build Solutions AI
MANDATE 02

The Tracker

Program Management & Execution
Performance tracking. Run the pipeline. Measure time saved and cost impact. Hold the portfolio dashboard. Prove it's working.
What the JD asks
Lead end-to-end delivery of AI-enabled client projects from concept through implementation
Develop frameworks to track AI performance: time saved, cost impact, and quality metrics
Maintain a portfolio dashboard highlighting adoption, client outcomes, and return on innovation
What I have done
Portfolio-level tracking, built in
AESOP tracks cost, status, and outcomes across every agent in a program. Operationalization layer produces a RACI matrix, adoption plan, and maintenance plan per build — before launch, not after.
AESOP · Legal AI OS
ROI framework with defensible numbers
Cost impact = time saved × the right rate card, by practice group and client. Calibrated baselines with documented methodology and confidence levels. Quality tracked by override rate, not satisfaction surveys.
ROI · Cost Impact · Quality Metrics
ROI and time-to-value as a discipline
Cut enterprise time-to-value from 6 months to 2 at Sopheon. 60% faster client response at Risilience. $320K ARR generated through consultative program design.
Sopheon · Risilience · Sigmaxim
MANDATE 03

The Teacher

Change Management & Enablement
Enablement and adoption. Build AI literacy. Grow the champion network. Close the gap between use and governance. Make the tools stick.
What the JD asks
Partner with L&D to upskill lawyers and staff in AI literacy and practical use
Build adoption programs that integrate AI into daily legal workflows
Promote a culture of experimentation, measurement, and continuous improvement
What I have done
Enablement at a scale most programs never reach
Weekly live instruction for cohorts of 150–200 practitioners at AI Build Lab (top-5 Maven program). 1,000+ member learning community. Curriculum design, assessments, governance frameworks, and responsible-deployment practices.
AI Build Lab · 150+ professionals monthly
Adoption inside a live enterprise
Ran AI literacy workshops and practitioner working groups across 4 departments. Enablement kit ships with the agent: workshop deck, prompt engineering guide for lawyers, adoption playbook, AI literacy FAQ. Departments with zero AI familiarity building their own agents within 3 months.
Enterprise engagement · Enablement kit
Learning infrastructure that scales
Built a 25-course sustainability academy with automated assessment at Risilience — 70% efficiency gain in training delivery, 35% increase in completion rates. $175K qualified pipeline from AI-powered knowledge platforms.
Risilience · Sopheon · Sigmaxim
MANDATE 04

The Protector

Governance & Risk
Responsible AI. Align to governance and ethics standards. Verify the confidentiality boundary actually holds. Prove it to clients and regulators.
What the JD asks
Ensure AI solutions align with the firm's governance, ethics, and risk standards
Respond to client RFPs and surveys on AI governance and data handling
Uphold confidentiality across all AI touching privileged client work
Strongest quadrant
For a firm that treats client confidentiality as existential, this is the mandate that matters most. The evidence below is the deepest I have.
What I have done
Audited the confidentiality wall — it held
Verified the access-control boundary between employment legal and commercial legal at a Fortune 500 SaaS — the exact confidentiality gap that had blocked adoption of prior contract-management tools. Tenant data isolation enforced at the database level, not in application code. The architecture guarantees what policy can only promise.
Enterprise engagement · Tenant isolation
Governance engineered into the build
AESOP 4-dimension scoring engine with hard veto power — Safety, Bias, Accuracy, Compliance. 990 tests. Any critical failure blocks deployment. Independent Harvey evaluation: same engine ported to Legal AI OS. Vendor can't grade its own homework. ABA Formal Opinion 512: 6 duties baked into compliance evaluation prompt, confidentiality weighted at 30% of the safety dimension.
AESOP · Legal AI OS · Harvey evaluation
ISO 27001 — Vanta-Certified Auditor, Passed First Try
Led preparation for ISO 27001 certification at Risilience and achieved it while there. Ran the internal audit. Vanta-certified auditor. This is governance at the organizational level — the same framework a law firm's clients demand evidence of in every RFP and security questionnaire. HIPAA-compliant voice AI architected at Build Solutions AI.
Audit trail and RFP response capability
Regulatory mapping mode pre-populates compliance templates with cited evidence from evaluation history. Append-only audit trail — every evaluation, every override, every drift alert timestamped and immutable. Governance packs for RFPs: ABA 512 compliance evidence, data handling documentation, one-click export.
Compliance · Audit · RFP response