AWS re:Invent 2026
Meet OnStak
at AWS re:Invent 2026
AWS is building what's next.
We get you there and keep the engines running.
Every AI has its villains: shadow AI, runaway tokens, blind spots no single tool sees. Want to meet the rogues quietly draining your estate?
See the villains →Strategic AI consulting and cloud engineering for the workloads that can't fail, run with AI after.
Three ramps onto the same track, depending on where you are. Move to the cloud, run it well, or bring us the specialist work. AWS is our lead cloud alliance.
Move to the cloud
- Cloud assessment & TCOWorkload profiling · FinOps model
- Migration readinessDependency mapping · wave planning
- Cloud migration & modernizationVMware, Windows, databases, containers
Run it well
- AI-powered operationsPredict, automate, self-optimize
- Capacity & FinOpsRightsizing · cost attribution by workload
- Managed operationsRunbooks, shadow ops, handover to your team
The specialist work
- HPC on AWSCompute farms, schedulers, license-aware capacity
- Silicon & EDAChip-design workloads, tape-out
- AI infrastructureGPU, inference, storage at scale
The AI Villains
Meet the Hero we built for you
AI Correlation Fabric: one screen showing all your tools, data and processes. So you can see what changed, what broke, and why. Instantly.
Connect with us at AWS re:Invent
Tue 1 Dec · 3:30 – 5:00 PM
Spritz Restaurant & Bar, The Venetian
An invitation-only hour with NetApp, off the show floor, with peers and the engineers behind the work, over a drink.
Add me to the guest list →1:2:1 demos
Tue 1 Dec – Thu 3 Dec
Caesars Palace, Las Vegas
A private viewing of AI Correlation Fabric on HPC with our cloud experts.
Book a private discussion →Thank you!
We have your request. Robert De Olivera will confirm a time with you shortly. A confirmation is on its way to your inbox.
Six of us on-site, and we'd love to meet you.
These are the people who do the work, and they are looking forward to meeting you. Real conversations, not pitches. Tell us what you are working on and we will make sure the right person is there.
Mike Thomas
My experience at AWS gives me insights and relationships to help our AWS cloud practice deliver higher business impact to our joint customers.
Ahmet Hussein
The architecture call before you commit: HPC, silicon and AI infrastructure on AWS.
Ben Haddox
The brain behind AI Correlation Fabric, and how it can enable optimization from Day 0 through to Day 2.
Robert DeOliveira
Delivering the insight and technical expertise to drive value for our customers in the cloud.
Muhammad Haq
OnStak is human first. We built our business around caring about what our customers need. I want to know what matters to you.
Ahmad Fawad
How we work alongside your team · Structuring a first piece of work · Who to bring in, and when
Backed by our AWS partner standing.
AWS Advanced Tier Services Partner
With Migration Competency and Migration & Modernization Services, delivering to AWS frameworks and the Migration Acceleration Program.
ProServe partner in high tech and semiconductor
We land the heavy-consumption workloads AWS field teams can't take on alone, and stay for Day 2 so the consumption compounds. Specialist depth where it is scarce.
A senior team that knows the AWS ecosystem
A number of our senior engineers have spent much of their careers in and around AWS. Less time spent explaining context, more spent on the actual workload.
Across the line, and still running.
Delivered and audited, across public sector, semiconductor, healthcare and retail. Anonymized here, customer names available on request.
Lower annual run cost
A US state health agency: 20-plus systems migrated and modernized, 7,773 vulnerabilities remediated, near-zero downtime.
Servers migrated
A government services program moved in coordinated waves, around 60% faster, more than 2 petabytes of data.
Live audit passed
Regulated federal tax data migrated and defended across 13 IRS Safeguards sessions, without a finding.
Chip design never stopped
A global semiconductor maker moved a live silicon-design environment to AWS: 546 engineers, 95 EDA tools, around 200 TB, with no interruption to chip development.
Fewer severe fall injuries
A US healthcare network: patient-safety video analytics across 250-plus cameras.
More efficient
A quick-service restaurant chain: drive-thru video and voice AI across 500-plus branches.
Faster to build, cheaper to run.
AICF, our correlation fabric (patent-pending), is the power under the hood, and the correlated context that tells you where you are actually losing time. It is the AI we use to deliver and to run operations. Figures from our own instrumented testing, and we can demonstrate the token reduction live.
Fewer tokens per decision
For AIOps use cases, with a 2 to 3x floor in smaller environments. Lower AI cost, not just a lower cloud bill.
Months to modernize
Application modernization on our own estate, 30 to 40% less effort, using AICF.
Devices seen in one hour
A regional health system got a live view of every device calling an AI service: 1,122 sessions across 340 devices, in a single hour.
re:Invent runs on five themes. We lead on four.
Enterprise and agentic AI, cloud migration and modernization, data as the AI foundation, cost and efficiency, and security and resilience. Here is where OnStak lands in each.
We put AI into production, not pilots.
Pilots don't fail on the model, they fail on the infrastructure. We do the cloud work that gets AI running for real on AWS, and we stay to run it.
The workloads AWS field teams can't take on alone.
Silicon and HPC, risk, genomics, regulated data. We land them on AWS, pull the sustained HPC, GPU, inference and storage consumption, and speed up ProServe delivery.
Our own AI makes it faster and cheaper.
AICF correlates and certifies your data so AI runs on the real picture of your estate, with fewer tokens per decision. We accelerate time to value, and control what the AI costs.
Governed, and we run it after.
Day 2 is the practice: operational observability, FinOps, and governed AI you can run in your own VPC on AWS, with approval before any action and an audit trail for every step.
Book time with us at re:Invent
Grab a private 1:1 discussion, or a place at the NetApp reception. Tell us where you are stuck and we will bring the person who has done it.
Questions we hear most
What is OnStak doing at AWS re:Invent 2026?
OnStak is meeting customers and partners at AWS re:Invent 2026 in Las Vegas, from November 30 to December 4. Five of our people will be on site to talk about cloud migration, HPC, AI infrastructure and running AI safely. Book a meeting and we will make sure the right person is there.
What is AI Correlation Fabric (AICF)?
AI Correlation Fabric (AICF) is OnStak's patent-pending engine. It pulls compute, GPUs, schedulers, licenses, storage and cost into one correlated picture, profiles every job, and correlates and certifies the data before any AI sees it. It gives your team one screen to see what changed, what broke and why.
How does OnStak keep AI automation safe?
AI proposes, people approve, and code enforces. That rule keeps a person in control of every change. Compliance controls and evidence are built in.
Do we need Terraform or deep AWS skills?
No. Your team works in three steps: Ask, Review, Approve. You do not need Terraform or deep AWS skills to use it.
Do we have to replace our schedulers, licenses or storage?
No. There is no rip-and-replace. AICF works with the schedulers, licenses and storage you already run.
How do we get started?
Start with a two-week, read-only readiness assessment. The map and plan we produce are yours to keep. If you want to go further, a proof of concept on your workloads takes about four weeks.
Which compliance programs does it support?
It supports SOC 2, ISO 27001, HIPAA, GxP, ITAR and EAR. We provide the controls and evidence, and certification stays with the customer.
Is OnStak an AWS partner?
Yes. OnStak is an AWS Advanced Tier Services Partner, with the Migration Competency and ProServe delivery.
Trusted By
