Cloud Integrity Group · Architecture · Cost · Storage · Responsible AI

Senior technical ownership.
Architecture, spend, and AI that will hold.

I help companies on AWS stay technically healthy: architecture that will hold, spend that is defensible, storage in the right class, AI that is used responsibly, and support work that does not live in one person's head.

Offers

Three ways in

Most first engagements are a written review. If the work is clear after that, we either build the AI system or keep a fractional TAM seat.

A· 2 weeks · fixed fee

Architecture, cost, storage & AI review

A TAM-style pass over the environment. You get a written brief: what is fragile, what is overpaying, whether storage is in the right class, whether AI is being used safely, and the next 90 days in order.

  • Well-Architected style findings, prioritized
  • Cost optimization with owners and savings estimates
  • S3 tiers, lifecycle policies, Intelligent-Tiering
  • AI responsibility: data, access, evals, human handoff

B· 60–90 days · project

Applied AI support system

Stand up a voice or chat agent on your real content — product docs, runbooks, customer FAQs — so tier-1 questions stop landing on humans.

  • Knowledge base from existing sources
  • Voice + chat agents, handoff rules
  • Pilot you can put in front of leadership

C· monthly retainer

Fractional technical owner

A designated technical owner for companies that need TAM-style rigor without a full-time hire, or vendors that need a senior technical customer-success lead.

  • Health reviews and exec-ready updates
  • Escalation judgment and architecture guidance
  • Adoption plans for new AWS / AI services

Review modules

What the written brief actually covers

Buy one module or all four in the same two-week window. Same format every time: finding, risk, recommendation, owner, effort.

Architecture

Best-practice review

Reliability, security posture, networking, identity, data stores, and operational ownership. Not a recertification exam — a list of the few changes that reduce real incident risk.

Cost

Optimization review

Rightsizing, idle and orphaned spend, storage classes, commitment strategy, and “this is expensive because the architecture is.” Savings ranked by effort vs. dollars.

AI responsibility

Responsible-use review

Scored against the AWS responsible AI dimensions: fairness, explainability, privacy and security, safety, controllability, veracity and robustness, governance, and transparency.

Storage

S3 and data-tier review

Storage class mix, Intelligent-Tiering, lifecycle and retention policies, incomplete multipart uploads, replication, and access patterns. Recommendations ranked by dollars vs. retrieval risk — not a default “move everything to Glacier.”

Patterns

What the numbers usually look like

Illustrative mixes from TAM, cost, and AI-readiness reviews — the shape of the work, not a promised result. Your environment will differ; the brief is to put your actual numbers on this page.

Cost review

Where the bill usually leaks

  • Oversized compute28%
  • Idle resources22%
  • No commitments20%
  • Storage / snaps18%
  • Data transfer12%

AI support

Tickets an agent can actually take

  • How-to / FAQ38%
  • Needs a human25%
  • Status checks22%
  • Access & billing15%

Responsible AI

Typical first-pass gaps

  • Governance30
  • Controllability35
  • Explainability38
  • Veracity40
  • Fairness42
  • Transparency45
  • Safety48
  • Privacy & security55

Cost mix is share of recoverable waste, not share of the whole bill. AI scores are 0–100 against the AWS responsible-AI dimensions on a first look — governance and human control usually lag the model itself.

Fit

This is a good match if…

You run product or infrastructure on AWS

You are past “we lifted some servers” and into reliability, cost, security, and service adoption. You want someone who has already run architecture and cost reviews with enterprise customers.

Your support org is drowning in repeat questions

Customers or internal teams ask the same things. The answers exist on a site or in Confluence. You want an agent that uses that content and knows when to escalate.

You sell infrastructure or cloud software

Key accounts need a designated technical owner: health, renewals, escalations, and a clean story for executives.

You need the translation layer

Engineers speak tickets. Executives speak risk and roadmap. Architecture discussions on Tuesday, C-suite reviews on Thursday.

Method

Same cadence as a designated technical owner.

Clear owner. Written recommendations. Escalations handled in the open. No 40-page strategy decks that nobody implements.

  1. 1

    Fit call

    20 minutes. Is the problem real, am I the right person, and what does done look like.

  2. 2

    One-page scope

    Outcome, access needed, timeline, price, and what is out of bounds. Signed before work starts.

  3. 3

    Working sessions

    Weekly technical time plus a short written status: health, risks, decisions, next actions.

  4. 4

    Handoff

    Docs, owners, and a 30-day check so the system does not depend on me staying forever.

Lee Denning

Lee Denning

Lehi, Utah · dleedenning@gmail.com

About

Enterprise TAM who still opens the terminal.

Experience as a Technical Account Manager II at AWS, working with enterprise customers on cost optimization, cloud infrastructure, security, storage, and AI/ML workloads. Before that, designated TAM for VxRail at Dell EMC, after years on Isilon/PowerScale support.

The through-line is ownership: keep the environment healthy, get the right people out of incidents, and turn tribal knowledge into a process someone else can run.

AWS Solutions Architect ProDeveloper AssociateML Engineer AssociateAI PractitionerNVIDIA AI InfrastructureVMware VCP

Contact

Tell me what is on fire.

If it is a fit, you get a short reply and a time to talk. If it is not, I will say so quickly.

Prefer email: dleedenning@gmail.com·801-688-7733

Independent practice in formation