BH

Brian Himanek

Now building: agentic AI

Enterprise Growth and AI Commercialization Executive

I lead growth for strategic enterprise accounts, and I build AI systems to solve the problems I see firsthand.

Across AWS, Amazon, Deloitte, DXC and Iron Mountain I have run a 25-person revenue organization with a full P&L, built an enterprise team from zero and grown Fortune 500 portfolios. I also hold an M.S. in AI and machine learning, and I build the systems I recommend.

AWS

Amazon

Deloitte

DXC Technology

Iron Mountain

Brian Himanek

Tampa Bay, Florida

12
12
12

strategic enterprise accounts in my current portfolio

$42M
$42M
$42M

manufacturing services portfolio led at DXC

$17.5M
$17.5M

Fortune 500 portfolio at AWS, grown from $13.8M

$17.5M

Fortune 500 portfolio at AWS, grown from $13.8M

25
25

person organization led with full P&L

25

person organization led with full P&L

  • Strategic accounts

  • Agentic AI

  • Value engineering

  • P&L leadership

  • Manufacturing and industrials

  • Local LLMs

  • Executive sponsorship

  • Cloud, data and ML

  • Program rescue

  • Team building

Results

What happened when I owned the number.

Amazon Web Services

Start · $13.8M

End · $17.5M

$13.8M to $17.5M, at 127% of plan

Grew a portfolio of North American Fortune 500 accounts by tying cloud, data and machine learning adoption to each customer’s economics. Net revenue retention reached 118%.

DXC Technology

Value case

Co-authored with VP Ops

Phase 1 funded

A $15M transformation case, funded into Phase 1

Co-authored a three-year value case with a VP of Operations at a global manufacturer, built on productivity, cost and operating risk. Generated about $850K in qualified pipeline in six months across a $42M portfolio.

Phoenix Data

Before · 78% retention

After · 92% retention

Retention lifted from 78% to 92%

Led a 25-person revenue organization with full P&L across a $50M+ portfolio. The team delivered 150% of renewal and expansion quota four years running.

Deloitte

Stalled

Go-live

A stalled ERP program, taken to go-live

Took over an implementation that had stopped moving and helped get it across the line to go-live.

AI

I build AI agents that end in an operating result.

I hold an M.S. in AI and machine learning, and I put it to work. I design agent systems, run them on a live book of accounts, and hold them to the same standard as any other investment: a number has to move.

8

coordinated agent layers

12

living account files as shared memory

3

frontier models in the bake-off

128K

token context on local hardware

How I get AI to an outcome

01

Pick by the KPI

Start from the number the business needs to move, then choose the use case.

02

Fund it with a value case

Build the case with the operations and finance leaders, so the numbers are theirs.

03

Give adoption an owner

Data readiness, governance and the operating model get named owners before launch.

04

Track until it moves

Hold a steady cadence and measure value until the operating number changes.

Three systems I designed, built and run

Command center

Signal monitor

Loose ends

Offense engine

Builder agent

Red team

Board review

Personal rail

Memory spine · living account files

Operating system

An AI operating layer for a book of accounts

Eight coordinated layers on a memory spine of living account files: a weekday monitor that scans public signals and ranks the top five, a loose-ends tracker and a red-team seat. A person approves anything that reaches a customer.

Agentic workflows

Daily signal digest

Human approval gates

Why I built it

So each day starts from ranked signals and tracked loose ends.

Inside the operating layer: one loop, every weekday

01 Sense

Scheduled agents scan public signals on every account each weekday morning and rank the top five.

02 Think

A reasoning model weighs each signal against the account’s living file and proposes the next move.

03 Challenge

A red-team seat argues against the plan before any time is spent on it.

04 Build

A builder agent drafts the brief, deck, workbook or outreach.

05 Approve

A person approves anything that reaches a customer or the system of record.

Memory spine · one living file per account · every layer reads it and writes back

Two model families, split by job: scheduled cloud agents do the sensing, and a reasoning model does the thinking and building. A Friday board review closes the week.

01 Research

forensic brief

Gate

graded sources

02 Execution

field playbook

Field system

Two-stage account intelligence

A research stage that builds a forensic account brief with source-independence rules and graded provenance, then an execution stage that turns it into a sequenced field plan. Hardened through a bake-off across three frontier models.

Research to playbook

Graded sources

Multi-model tested

Why I built it

So strategy rests on confirmed facts. The playbook stage is handed a list of claims it may not treat as established.

brain://local

LLM

128K context · 51.7 tok/s

Image

generate + edit

Video

15 s clips with audio

Voice

speech in, speech out

Watchdog

heat + memory

Personal build

BRAIN, a private AI workstation

A fully local AI stack on a single 24 GB GPU: chat with a 128K context window, image generation and editing, video, voice in and out, and a set of task-specific assistants. It manages its own GPU memory and heat and runs a regression set to catch quality drift.

Local LLMs

Image + video

Voice

Runs offline

Why I built it

So private material stays on my own hardware, and so I know first-hand what local models can and cannot do.

How I build agents

Rule 1

Memory before autonomy

Every agent works from one living file per account, so context outlasts the session.

Rule 2

Research stays neutral

Commercial assumptions are kept out of the research stage and enter later as labeled hypotheses.

Rule 3

Every claim is graded

Sources are graded, and anything recalled from model memory is marked for verification.

Rule 4

People approve what leaves

Nothing reaches a customer or a system of record without a person signing off.

Rule 5

Pilot against a baseline

A new agent runs beside the current process for a set window before it replaces anything.

Rule 6

Private stays local

Sensitive context stays on my own hardware. Cloud models get a sanitized brief.

How I work

Understand what is at risk, what is possible, and what it takes to turn a good idea into an operating result.

At AWS, I helped bring cloud, data and machine learning programs into production by connecting technology decisions to customer economics. At Deloitte, I took over a stalled ERP implementation and helped get it to go-live. At DXC, I led a $42M manufacturing portfolio and worked with an operations executive to shape a $15M transformation case around productivity, cost and operating risk.

Today I lead a cross-functional team across a portfolio of large manufacturing and industrial accounts. I have sat in both seats, running a 25-person organization with a P&L and carrying accounts myself, and I know what each one needs from the other.

Three things I hold to with every customer

Start with what is at risk

Executives fund what protects or grows the business. I learn the exposure and the upside before I talk about a solution.

Tie technology to economics

A program earns its budget through a value case the CFO and the operations leader both recognize as their own numbers.

Stay until it runs

A signed deal is a starting point. The result counts when the customer’s people use it and the operating numbers move.

Career

Where I have done the work.

Iron Mountain

2026 to present

Client Partner Leader, Global Industries & Strategic Verticals

Own strategy and executive relationships for a 12-account North American portfolio across manufacturing, distribution and business services.

DXC Technology

2025 to 2026

Consulting Partner, Consulting & Engineering Services

Led a $42M manufacturing portfolio and shaped a $15M transformation case with an operations executive.

Stratus Group

2022 to 2024

Vice President, Strategic Accounts

Strategic account and AI transformation advisory at a boutique firm.

Amazon Web Services

2020 to 2022

Account Partner, AWS Professional Services

Grew a North American Fortune 500 portfolio from $13.8M to $17.5M and brought cloud, data and machine learning programs into production.

Deloitte

2016 to 2020

Senior Manager, Digital & AI Solutions

Took over a stalled ERP implementation and got it to go-live, and led a Salesforce transformation the client expanded into a broader engagement.

Amazon

2011 to 2016

Senior Manager, B2B Client Solutions

Built and coached the five-person client-solutions team behind Amazon’s enterprise B2B procurement offering, from early launch through Fortune 500 adoption.

Phoenix Data

2005 to 2011

Vice President of Account Management

Ran a 25-person account-management and revenue organization with the P&L for a recurring-revenue SaaS portfolio.

STX Technologies

2000 to 2005

Director, National Accounts

Owned national-account strategy, executive relationships, retention and expansion for enterprise technology customers.

Education and credentials

Education

M.S., Artificial Intelligence and Machine Learning

Colorado State University Global, Distinguished Scholar

MBA

West Texas A&M University

B.S., Business IT Management

Western Governors University

Certifications

CompTIA Security+

Certified Ethical Hacker (CEH)

Computer Hacking Forensic Investigator (CHFI)

ITIL Practitioner

Infor EAM

FranklinCovey Strikingly Different Selling

Contact

Let’s compare notes.

Enterprise growth, applied AI, or the place where the two meet.

Based in

Tampa Bay, Florida

© 2026 Brian Himanek

Designed in Tampa Bay. Built with AI, naturally.