I've spent more than a decade building software across distributed systems, cloud infrastructure, enterprise platforms, identity, cybersecurity, financial systems, integrations, developer tooling, and AI.

I currently work as a Lead Software Engineer at EPAM Systems, focused on enterprise modernization, architecture, technical direction, validation, and AI-accelerated software engineering.

I'm also the founder of Fenix Alliance and the architect behind Alliance Business Suite, a long-running attempt to answer a deceptively difficult question:

What would enterprise software look like if the business operated as one coherent system instead of a collection of disconnected applications?

That question has taken me from CRUD applications to distributed execution, payments, identity, digital trust, durable workflows, multi-tenancy, cloud infrastructure, AI agents, authorization, reconciliation, and runtime governance.

These days, a surprising amount of my job looks like managing fleets of software agents.

I delegate labor, not judgment.

Alliance Business Suite is a modular, multi-tenant platform for composing business capabilities into applications, workflows, integrations, and intelligent operating surfaces.

The interesting part isn't how many modules it contains.

The interesting part is that they are expected to obey the same laws.

Identity, money, documents, workflows, commerce, AI, integrations, authorization, and execution should not become separate universes simply because different teams or interfaces use them.

A business operation should remain the same operation whether it is initiated by a person, an API, a workflow, or an AI agent.

flowchart LR
    H["People"]
    UI["Applications"]
    AI["AI Agents"]
    WF["Workflows"]
    API["APIs / SDKs"]

    CAP["Governed Business Capabilities"]

    EXEC["Durable Execution"]
    AUTH["Identity & Authority"]
    DATA["Business Model"]
    EVID["Evidence & Observability"]

    H --> CAP
    UI --> CAP
    AI --> CAP
    WF --> CAP
    API --> CAP

    CAP --> EXEC
    CAP --> AUTH
    CAP --> DATA
    CAP --> EVID

Many ways to operate the business. One operating model underneath.

I'm much more interested in AI as an execution and reasoning substrate than as a chatbot bolted onto existing software.

The model I'm working toward is:

flowchart LR
    HUMAN["People decide"] --> AI["AI reasons"]
    AI --> WF["Workflows coordinate"]
    WF --> APPROVAL["Authority is established"]
    APPROVAL --> CAP["Capabilities execute"]
    CAP --> EVIDENCE["Evidence proves"]

    POLICY["Policy"] -.-> AI
    POLICY -.-> WF
    POLICY -.-> APPROVAL
    POLICY -.-> CAP

That leads to problems I find far more interesting than prompt engineering:

execution identity · authorization · durable approval · idempotency · leases · fencing · recovery · provenance · tool governance · provider abstraction · human oversight

An agent being capable of doing something does not mean it should be allowed to do it.

And an AI-generated answer being plausible does not mean it is correct.

The same business capabilities can be exposed through different experiences without duplicating the business rules behind them.

The interface changes. The business operation should not.

Frameworks are temporary. Business concepts tend to survive them.

Distributed systems become interesting when the network times out after the other side already succeeded.

Identity, tenant scope, policy, authorization, approval, and execution context should travel with an operation.

If UI, workflows, APIs, and agents perform the same business action, they should converge on the same capability.

If a system cannot explain what happened, operating it becomes archaeology.

Cheap implementation makes architecture, verification, coordination, and change-impact reasoning more valuable — not less.

Complexity is acceptable when it removes a harder recurring problem.

A modular Business Operating System for composing and governing enterprise capabilities.

Cloud infrastructure and managed services around ABS.

Multi-tenant commerce and procurement infrastructure.

Technology education and professional learning.

Software engineering, cloud, AI, data, and security services.

Large-scale modernization. Distributed systems. AI-native platforms. Identity. Financial infrastructure. Developer platforms. Multi-tenant architecture. Cloud infrastructure. Cybersecurity.

But the common thread is simpler:

I like taking systems that are complicated for the wrong reasons and making their underlying model clear.

Motorcycles · Formula 1 · Running · Boxing · Tennis · Aviation · History · Strategy

I have a habit of getting interested in systems regardless of whether they're made of software, people, engines, markets, aircraft, or civilizations.