NVIDIA Inception Member · Active Client Engagements

IT consulting that
runs on its own AI.

We build and operate our own AI platform — and we deliver the same capability to our clients.

A decade of AI engineering. Deep Azure expertise. Infrastructure residencies delivered under Dell and Unisys. NVIDIA Inception member.

Credentials
NVIDIA Inception
Dell Engagement
Unisys Residency
Microsoft Azure & AKS
Azure Local & Hyper-V

Four Proven Offerings

We only advertise what we've delivered. Every offering is backed by a real engagement.

🏗️

AI-Augmented Infrastructure Residencies

A senior principal embedded for defined weeks, backed by an AI delivery workforce. Every hour produces durable artifacts — designs, runbooks, validated configs — not just advisory output.

Multi-site Hyper-V & SCVMM residency, state-government end client — delivered under a Dell residency contract (Unisys / Kansas, 2026)
☁️

Azure & Hybrid Platform Engineering

Fixed-scope design, build, rebuild, and migration of Microsoft virtualization and hybrid platforms — Azure Local, Hyper-V FC, SCVMM, Network ATC. Dry-run-safe automation with per-phase rollback plans baked in.

KSU two-node cluster design v1.1 (revised against customer engineering review) + Santa Barbara Azure Local clean-rebuild SOW
🔄

Application Migration Factory

Dell's 3-phase methodology operationalized as a governed migration factory: phase gates, generated runbooks and rollback plans, human go/no-go checkpoints, and a glass-room oversight console.

Productized in active Dell engagement (B2B2B channel delivery)
🤖

AI Workforce Enablement

We run our own consulting business on an AI workforce with principal oversight — we stand up the same model in client delivery organizations. Not a framework recommendation. A working production system we dog-food daily.

Luca platform in production on real contracted work; NVIDIA Inception credentials + active Dell / Microsoft stack depth

How We Work

Every engagement starts with a no-obligation 30-minute infrastructure assessment. Scoped per engagement — no hourly rate cards on the website.

01

Residency

Multi-week hour blocks. A named principal on-site or remote, AI workforce behind them, weekly artifact delivery.

02

Fixed-Scope Design Sprint

2–4 weeks. Assessment, architecture, and a decision-ready design document with a go/no-go recommendation.

03

Design–Build

SOW-defined scope, fixed outcome. We design, build, and hand over. Used for Azure Local rebuilds and migration factories.

04

Channel / Partner Delivery

We deliver under primes — both of our completed engagements are B2B2B. If you're a systems integrator, that's a strength, not a footnote.

Luca AI — the platform our AI workforce runs on.

Not a vendor recommendation. We built it, we operate it, and we run real contracted work through it every day. We deploy the same capability for clients who want it.

luca@ai ~ check infrastructure status
✓ VMs monitored across Azure, AKS, Hyper-V
✓ Services healthy · 0 incidents
✓ Active delivery engagements: 2
luca@ai ~ deploy service to production
✓ Building container image via ACR...
✓ Deployed · zero downtime · rollback ready
🧠

AI Engineering Workforce

Autonomous engineering personas (architect, builder, tester, deployer) operate under named-principal oversight — the same model we embed in client orgs.

📧

Integrated Business Operations

CRM, communications, project tracking, calendar, billing — all connected. Our consultancy runs on it. A real reference install, not a demo environment.

🔌

Azure-Native by Design

Kubernetes-first on AKS, Azure Container Registry, CloudNativePG, Cloudflare tunnel — the same stack we build for enterprise clients.

🔒

Enterprise Security Architecture

RBAC with resource-level permissions, audit logging, AD integration, zero-trust network posture — built to the standards we consult on.

Luca AI is also available as a standalone platform for organizations that want to run their own AI workforce.

lucaexpress.com →

A Claude-powered AI workforce, deployed on your cloud and run by your people.

We deploy and operate a workforce of role-based AI personas — PhD-level reasoning models and senior-engineer-grade coding models — wiring Claude, Claude Code, and the Anthropic API into the Luca Express ecosystem. Each persona runs like a real team member: it takes a ticket, opens a branch, writes and tests code, and ships through CI under your team's oversight. We are an independent integration and managed-service partner, not Anthropic: bring your own Anthropic subscription and API key, or have us manage them for you.

👨‍💻

AI Software Engineer

A programmer that turns a ticket into working software — it writes the module, the routes, the interface, and the tests, then opens a pull request for review. It works to your standards, not from a blank prompt.

A real operator, not a chatbot: it opens its own branch, runs the pre-deploy check, and ships a single squash-merge PR per task.
🐛

AI Bug Fixer

An engineer that reproduces a reported bug, traces it to the root cause in the owning module, and applies a minimal, reversible fix. It closes the ticket with a verification note so the fix is auditable.

Ticket-first discipline: it opens a ticket before any code, logs the root-cause note, fixes in the source module, tests, then closes — a continuous, backlog-draining sweep.
🔍

AI Code Reviewer

A reviewer that gates every change before merge — checking security, authentication, SQL safety, code quality, and UI standards against your platform contract. It is the discipline that keeps the trunk clean.

The queue gatekeeper: a structured PASS/FAIL/WARNING review on auth guards, parameterized SQL, and XSS escaping — low-risk changes merge only once its checks are green.
🧪

AI QA / Test Engineer

A QA engineer that writes and runs the tests protecting your trunk — health endpoints, API contracts, auth gates, and responsive UI on both desktop and mobile. It catches regressions before your users do.

Defends the merge queue with end-to-end tests; every missed bug that triggers a rollback gets a new regression test before the offending change can return.
📦

AI Migration Engineer

A migration engineer for the classic 'modernize without downtime' job — it moves legacy applications into a modern modular structure without breaking the surface your users depend on today.

Uses the Strangler Fig pattern: old endpoints keep responding at every step, changes ship as small revertible PRs, and old code is removed only after the new path has soaked in production.
🚀

AI Release / DevOps Engineer

An ops engineer that owns the path from merged code to a running deployment — pre-deploy checks, risk classification, build, rollout, verification, and a recorded release, with rollback on failure. It ships, then watches.

A strict, auditable Kubernetes-native pipeline: snapshot, risk-classify the diff, build, deploy, watch health and latency, record the release, and auto-roll-back on a smoke-test or SLO breach.

From assessment to a workforce your team operates.

01

Assess & Scope

We start with your environment, your repositories, and the work you want an AI workforce to own. We agree the personas, the guardrails, and the human-oversight model before any deployment.

02

Deploy on Your Cloud

We stand up the Luca Express orchestration layer on your Kubernetes — Azure Local, AWS, Google Cloud, or Anthropic direct. Your data and your deployment stay yours; we wire in Claude and the Anthropic API on your terms.

03

Stand Up the Personas

We provision the AI workers — each with its own dev station, Claude Code terminal, and identity — and connect them to your ticketing flow. They begin operating under presence-gated human oversight from day one.

04

Train & Hand Over

We train your team to direct, supervise, and take the wheel from the personas, then hand over operations — or keep running the workforce for you as a managed service. Your call.

Claude + Luca Express, run by your humans.

This is the model we operate ourselves: Claude, Claude Code, and the Anthropic API orchestrated through the Luca Express ecosystem, with role-based personas doing the work under continuous human control. The workforce we deploy for you is the one our own consultancy runs every day.

🖥️

Claude Code Dev Stations

Each AI worker runs on its own dev station — its own clone of your repository, its own Claude Code terminal, and its own identity — so it operates like a real member of the team, not a shared chat window. Work flows in through a ticketing system: one task, one branch, one pull request, CI, merge, deploy.

🌙

A Workforce That Compounds

Each persona runs a continuous cognitive loop — perceive, reason, plan, act, reflect — rather than a one-shot prompt, and its real work is captured as a structured dataset. While the fleet is idle, each engineering persona reviews its own work history and consolidates the experience, so the workforce learns on the job. Your data stays within your deployment.

🎛️

Personas Operate Claude — You Stay in Control

The personas operate the Claude console and the Anthropic API directly, on your behalf. A presence-gated takeover lets the responsible human step in and take the wheel at any moment. The principle is simple: personas do the work under human oversight, never instead of it.

🧩

Integrated into Luca Express

The workforce is orchestrated through the Luca Express ecosystem — the live platform our own consultancy runs on, not a slide deck. Ticketing, oversight, business operations, and the CD pipeline are connected, so the personas work inside a real, running system from the first day.

Deploy anywhere there is Kubernetes.

The workforce is cloud-agnostic and Kubernetes-native by design. The same orchestration runs on your Azure Local cluster, on AWS, on Google Cloud, or against Anthropic and Claude directly — wherever there is K8s. You keep your cloud, your data residency, and your security posture; we deploy into it and train your team to operate it.

Azure Local
AWS
Google Cloud
Anthropic / Claude direct
Any Kubernetes
Independent partner — not Anthropic. Claude and the Anthropic API are products of Anthropic PBC. Microsoft Azure, AWS, Google Cloud, NVIDIA, Dell and Unisys are trademarks of their respective owners. Gus IT LLC is an independent integration and managed-service partner and is not affiliated with, endorsed by, sponsored by, or owned by any of these companies. We couple their services to the Luca Express platform on your behalf — under your own accounts, your own subscriptions, and your control.

Transparent pricing — Claude billed by Anthropic, our service fee on top.

How billing works: You pay Anthropic directly for Claude. You bring your own Claude account or API key — your Pro, Max, Team, or Enterprise subscription is billed by Anthropic, at Anthropic's rate, on Anthropic's terms. Gus IT is an independent integration and operations partner; we do not resell Claude, and we do not set or control Anthropic's pricing. Separately, you sign a month-to-month service agreement with Gus IT for our work on top of Claude: the Luca Express / AIOS integration, the persona orchestration and day-to-day operations, deployment, and support. The headline number on each card is our service fee only — what we charge to run the platform around your Claude subscription. It is month-to-month: cancel any time, just like Claude. All prices are in USD and exclusive of VAT. Buyers in Germany and the EU self-account for VAT under the reverse-charge rules where applicable; Gus IT LLC is a U.S. entity and does not charge German VAT.
Assisted Seat
Pairs with Claude Pro (standard) — billed by Anthropic at their rate
+ $39 / mo
Gus IT service fee · month-to-month · per seat · excl. VAT
Your Claude, working inside your business — assisted, governed, supported.
  • Claude Pro is billed by Anthropic — you bring your own account or API key
  • Our fee covers the Luca Express / AIOS integration so Claude works with your data and tools
  • Single sign-on, role-based access, and your-data-only isolation we configure and operate
  • Hosted on our managed Kubernetes platform — compute, database, storage and networking included
  • Email support, platform updates, and ongoing maintenance
  • Month-to-month — cancel any time, just like Claude
Start an Assisted Seat
AI Worker
Pairs with Claude Max — billed by Anthropic at their rate (Max starts higher than Pro)
+ $149 / mo
Gus IT service fee · month-to-month · per seat · excl. VAT
An autonomous worker we orchestrate, supervise, and keep running for you.
  • Claude Max is billed by Anthropic — you bring your own account or API key for the higher-usage tier
  • Our fee covers the autonomous-worker orchestration: a persona that perceives, plans, acts, and reports
  • Persistent runtime — terminal, jobs and the cognitive loop kept healthy with active oversight tooling
  • Hosted on our managed Kubernetes platform — the heavier per-seat compute, database and storage share is included
  • Priority support, supervision dashboards, and ongoing maintenance
  • Month-to-month — cancel any time, just like Claude
Deploy an AI Worker
API / Pay-as-you-go
Claude API usage — billed by Anthropic at their per-token rate, directly to you
Usage + 18% (min. $99 / mo)
Gus IT management margin on routed usage · month-to-month · excl. VAT
Pay Anthropic for what you use; pay us a transparent margin to route and manage it.
  • Claude API usage is billed by Anthropic, per token, directly on your own key
  • Our margin is a transparent percentage on the usage we route through the platform — no hidden markup
  • Cost levers we apply for you where it fits: batching, prompt caching and model-tier routing
  • Hosted on our managed Kubernetes platform — request routing, the model bridge, and usage metering
  • Usage dashboards so the Anthropic line and our margin are always separated and visible
  • Month-to-month — cancel any time
Get an API estimate
Enterprise
Pairs with Claude Enterprise — seats and usage billed by Anthropic under your agreement with them
Custom
Gus IT service fee · scoped to seats & scope · excl. VAT
Run it on your cloud, on your terms — we deploy, integrate, train and operate.
  • Your Claude Enterprise agreement is held with Anthropic — seats and usage billed by them
  • Deploy on your own cloud or in our managed environment — your choice of substrate
  • Initial deployment, integration and team training scoped and quoted up front
  • Our service fee is scoped to your seats and scope, with governance, SSO/SCIM and audit support
  • Dedicated operations, security review, and a named delivery contact
  • Flexible term — we agree the commitment that fits your governance
Contact sales
$250/hour

Deployment, training and advisory at $250/hr (excl. VAT). For initial integration, onboarding your team, architecture and IT-governance advisory, or a one-off project — billed by the hour, no subscription required. This is our work; any Claude usage during the engagement is billed to you by Anthropic on your own account.

Book a 30-min Assessment →

See it in practice — our Use Case library.

We publish downloadable, PhD-level use cases — white papers, decks and HTML docs — drawn from real AI-workforce delivery, in English, French, German, Portuguese and Spanish. New cases are added continuously.

Browse the Use Case library →

Built by practitioners, for practitioners.

Gus IT is a US-registered IT consulting firm serving DACH clients. We combine a decade of hands-on AI engineering with deep Microsoft infrastructure expertise — and we use our own AI platform to deliver.

We deliver under primes. Our channel partners are Dell and Unisys. Our credentials include NVIDIA Inception membership and a delivered residency for a state-government end client.

The AI-Augmented Delivery Model

Every engagement is led by a named human principal accountable for outcomes. Behind them, an AI workforce handles documentation, runbook generation, configuration review, and progress reporting — freeing the principal to focus on the architecture decisions that actually matter.

G
Gustavo Assunção
Founder & Principal Engineer

A decade building and operating AI systems in production. Deep specialization in Microsoft Azure, AKS, Hyper-V, and hybrid infrastructure. Led the architecture and delivery of Luca AI from concept to active client deployments. Engagements with Dell and Unisys on enterprise infrastructure residencies. NVIDIA Inception member.

Azure & AKS Hyper-V / Azure Local AI Systems NVIDIA Inception Dell Channel Unisys

Book a 30-minute infrastructure assessment.

No obligation. We look at your current environment, identify the highest-impact opportunities, and tell you honestly whether we're the right fit.

📞

Phone

DE: answered by our AI platform — routes to the right person.

(Numbers provisioning in progress — use email for fastest response)

📍

Legal Entity

Gus IT LLC · Registered in Florida, USA

Serving DACH clients remotely and on-site

Request an Assessment
We respond within one business day.

Your information is used only to respond to your request.

Impressum

Legal Entity

Gus IT LLC
Registered in Florida, USA
Representative: Gustavo Assunção

Contact

E-Mail: [email protected]
Web: gusit.de

EU Representative (Art. 27 GDPR)

Gus IT LLC is a US-registered entity targeting EU clients. An EU representative designation under Art. 27 GDPR is under review.

Liability for Content

The contents of these pages were created with due care. However, Gus IT LLC assumes no liability for the accuracy, completeness, or timeliness of the information provided.

Trademarks & Affiliation

Claude and the Anthropic API are products of Anthropic PBC. Microsoft, Azure, AWS, Google Cloud, Kubernetes, NVIDIA, Dell, Unisys and all other product and company names are trademarks or registered trademarks of their respective owners. Gus IT LLC is an independent integration and managed-service partner and is not affiliated with, endorsed by, sponsored by, or owned by Anthropic or any other named company. Where we deploy these services, we couple them to the Luca Express platform on the customer's behalf — under the customer's own accounts and subscriptions. References to these names describe interoperability and our memberships, engagements, or delivered work, and do not imply any endorsement.

⚠ Legal review pending: A formal Datenschutzerklärung (privacy policy) will be published separately. If you have a data-related request, contact [email protected].