NF.Résumé
Nathan Frodsham

Nathan Frodsham

Strategy · Transformation · AI operations

I turn strategy, analytics and AI into systems teams can actually run.

I lead the work from business problem to operating model—aligning people, decisions and implementation across enterprise analytics, healthcare and AI operations.

Explore selected work Start a conversation
01Frame the real problem
02Design the operating model
03Enable accountable execution

Selected experience

ACCENTUREAMAZONOMNITURE / ADOBEOGILVY1-800 CONTACTSEVOLVE MEDICAL

Selected transformation work

The work, the decisions, and my contribution.

Five examples of strategy, analytics and AI becoming decisions, operating models and accountable execution. Open each case study for the constraints, decisions, ownership and role of AI.

01

Strategy & launch governance

Client work · pre-launch staging · 2026

Preparing a specialty-care practice for launch

A specialty-care WordPress site built from a 13-page Phase 1 scope, with a no-PHI intake design, clinical review gates and an operating handoff.

13-page Phase 1147-item clinical review logNo-PHI intake design
Read the case study

The problem

A new joint-care practice needed more than attractive pages. Its positioning, clinical language, referral boundaries, intake flow, ownership, and future acquisition system all had to agree before launch.

My contribution

I shaped the market position, scope and decision process; translated partner disagreements into bounded choices; designed the no-PHI architecture; and kept launch claims visibly gated until the clinicians approved them.

Important decisions

Practice ownership over platform lock-in; a focused joint-care identity over category sprawl; clinical-review markers over polished but unverified copy; and a site structure that can support SEO and paid acquisition without a rebuild.

Where AI helped

AI agents accelerated competitive research, copy drafting, implementation and QA. I wrote the constraints, made the strategic calls, reviewed the work, and remained accountable for the client result.

02

AI operating model

Operating company · automation since 2025 · multi-agent work since 2026

Designing accountable AI operations for a regulated brand

A connected research, content, monitoring and approval workflow that makes AI useful with defined human-review and publishing controls.

Approval-gatedCompliance-awareMulti-channel operations
Read the case study

The problem

A regulated wellness business faced severe channel constraints, a small team, and a constant need for fresh research and compliant communication.

My contribution

I designed the operating model across AI agents, workflow automation and private infrastructure: what the agents watch, what they draft, what evidence they retain, and where a human must approve.

Important decisions

Separate research from publishing; put compliance rules inside the workflow; route exceptions to a human; and favor useful, credible communication over high-volume automation.

Where AI helped

Models perform bounded research, classification and drafting. I designed the system, supplied the business and regulatory context, set the acceptance rules, and verify consequential outputs.

03

Enterprise decision support

Amazon · 2012–2013

Turning customer-service data into decisions at Amazon

Contact-driver analysis, operational KPIs and dashboards that supported resource allocation and VP-level decisions in global customer service.

Behavioral clusteringOperational KPIsVP-level decision support
Read the case study

The problem

A global customer-service organization needed to understand what drove customer contacts, identify recurring friction and decide where to focus operational attention.

My contribution

As Senior Business Analyst, I used clustering and behavioral analysis to surface patterns in customer contacts, built dashboards and operational KPIs, and provided decision support to VP-level stakeholders.

Important decisions

Connected analysis of contact drivers to resource-allocation questions, and paired detailed behavioral analysis with reporting that senior leaders could use.

Where AI helped

This work predates generative AI. It drew on statistical analysis, customer-service data and business context—the same foundation I bring to AI-enabled decisions today.

04

Analytics leadership

Evolve Medical · 2017–2021

Moving healthcare data from reporting into clinical preparation

EHR and appointment data translated into preventive-care candidate lists, pre-visit decision support and operational reporting for a growing medical practice.

EHR analyticsPredictive modelingClinical workflow
Read the case study

The problem

Useful opportunities were buried across clinical and operational data. The practice needed repeatable ways to identify candidates, prepare visits and make better decisions without asking clinicians to become analysts.

My contribution

As Director of Analytics, I designed the queries, models and reporting workflow; connected analysis to the people who could act on it; and supported broader digital and market strategy.

Important decisions

Operationally usable outputs over impressive models; clinician review over automated conclusions; and repeatable data products over one-off analysis.

Where AI helped

This work predates today’s agentic tools. It established the pattern I still use: combine domain context, rigorous analytics and a workflow that makes the result actionable.

05

Decision intelligence & risk

Self-initiated · private system · ongoing

Designing a systematic research and risk-control platform

A multi-workflow research system that turns market data into auditable decisions, with independent risk checks and monitored failure paths.

Multi-timeframe researchIndependent risk checksPaper-tested before live
Read the case study

The problem

Systematic investing fails when research logic, production data, execution rules and risk controls drift apart. The hard problem was not a single signal; it was maintaining a coherent decision system.

My contribution

I developed the strategy logic and operating architecture, set the experiment and validation standards, and coordinated repeated audits across data ingestion, regime classification, sizing, execution and monitoring.

Important decisions

Signals computed only from completed data; several independent safety checks rather than one all-in-one model; small, monitored rollouts rather than broad changes; and evidence logs rather than after-the-fact explanations.

Where AI helped

AI agents assist with code, diagnostics and adversarial review. I define hypotheses, approve changes, validate behavior against source data, and control every live-money boundary.

Experience

Strategic judgment, technical fluency, accountable execution.

2021—Now

Strategy, analytics & AI-enabled operations

Private therapeutic innovation firm · Independent initiatives

2017—2021

Healthcare & regulated-health analytics

Evolve Medical · Regulated health retail launch

2011—2017

Enterprise analytics & strategy

Amazon · Prophet · 1-800 Contacts · Omnicom · VCI

Earlier

Digital analytics, consulting & company building

Accenture · Omniture/Adobe · Ogilvy · Nu Skin · MTrove

How I work

AI is leverage. Accountability stays human.

I use AI agents as a research and implementation team—not as a substitute for domain judgment. I define the business problem, constraints, architecture, evidence standard and acceptance criteria. Then I review, test and own the result.

“Nathan helped my company grow… We recorded instant ROI impact using these personas in targeted campaigns.”— Tristan Webb, client recommendation on LinkedIn
  1. 01

    Clarify the decision

    What must change, who owns it, and what evidence would prove it worked?

  2. 02

    Design the controls

    What may AI do, what needs review, and where must the system fail closed?

  3. 03

    Verify the end state

    Test the output against primary evidence and leave a system others can run.

Transformation leadership · Analytics · AI operations

Working on a complex problem that needs to become a working system?

Based in Salt Lake City. Open to full-time leadership roles and fractional engagements in analytics, AI operations and transformation.

Email Nathan LinkedIn