Systems & Full-Stack Software Engineer combining a rigorous mathematical foundation (First-Class Honours across 1st & 2nd Year Computer Science; 86% in Formal Logic) with industrial placement experience engineering enterprise microservices at Next plc and multi-tenant data platforms at Valdris.

Focused on high-throughput backend systems, deterministic multi-agent software engineering (The Software Factory), quantitative & algorithmic modeling, and applied AI/computer vision.

C# / .NET · Blazor WebAssembly · Distributed Microservices · AI Attribution 2.0 · Enterprise Agentic Enablement

• Leading UI and client-state re-architecture for AI Attribution 2.0 using Blazor WebAssembly across a multi-repository, multi-solution enterprise microservice ecosystem.
• Engineering high-reliability cross-environment services, automated product classification workflows, and enterprise configuration pipelines within the Product Information & Management (PIM) Strategic Team.
• Enterprise Agentic Adoption & Developer Onboarding: Developing internal engineering playbooks and onboarding fellow developers into structured agentic workflows — codifying domain context (CONTEXT.md, ADRs), designing deep-module boundaries to contain blast radius, and enforcing deterministic verification gates across multi-repository services.

PostgreSQL Row-Level Security (RLS) · TypeScript · Anthropic Claude SDK · Deterministic Z-Score Engine · Playwright & Vitest

• Architected a multi-tenant sports science analytics and readiness platform enforcing strict database-level tenant isolation via PostgreSQL Row-Level Security (RLS) bound directly to cryptographic JWT claims.
• Engineered GDPR Article 9 dynamic special-category health data masking and a deterministic Composite Load Score (CLS) engine computing 28-day rolling z-scores ($\mu, \sigma$ windowing with cold-start dampening).
• Built an Anthropic Claude SDK compliance firewall with deterministic regex pre/post-interceptors to prevent diagnostic medical hallucinations, backed by automated cross-tenant security suites in Vitest and Playwright.

Python · Mathematical Benchmarking · LLM Reasoning Verification · Algorithmic Analysis

• Authored formal evaluation rubrics and executable Python verification suites to benchmark frontier LLM reasoning across discrete mathematics, data science, and algorithmic edge cases.

Architecture Blueprint & Case Study: Enterprise Agentic Systems: The Software Factory & The Lab (maciekgeneja.me/blog/enterprise-agentic-systems) ↗

Rather than treating AI coding as ad-hoc prompting, I engineer a deterministic Software Factory that allows engineering teams and autonomous agents to ship production software with zero architectural drift:

1. In-Repo Domain & Context Codification (CONTEXT.md · AGENTS.md · ADRs): Translating implicit tribal architecture into version-controlled ubiquitous language and Architectural Decision Records so newly onboarded engineers and autonomous agents reason from identical ground truth.
2. Deep Modules & Blast-Radius Containment: Structuring codebases around Ousterhout-style deep modules (narrow, strongly typed contracts hiding complex internals), static dependency boundary rules (dependency-cruiser), and isolated git worktrees so concurrent changes never collide or leak across service seams.
3. Composable Skill Workflows & Spec-Driven Execution: Standardizing delivery through modular execution loops (Stress-Test Grilling → Technical Specification → ADR → ready-for-agent Vertical Slices → Red-Green-Refactor TDD) gated by deterministic verification (npm run check, Playwright, Vitest).
4. Team Onboarding & Organizational Leverage: Upskilling developers to transition from single-turn chat prompts to structured multi-agent orchestration with strict human-in-the-loop architectural sign-off.

flowchart LR
    A[Requirement / Problem] --> B[Grill & Stress-Test]
    B --> C[CONTEXT.md + ADR Spec]
    C --> D[Isolated Worktree Slices]
    D --> E1[Subagent Slice A]
    D --> E2[Subagent Slice B]
    E1 --> F[Deterministic Gate: Types + Tests + Runtime]
    E2 --> F
    F --> G[Human Review & Merge]

Alongside enterprise repository architecture, I run the Software Factory across The Lab — the computers I already own linked over a private Tailscale and SSH network to run autonomous software development, research, and verification workflows on a placement-student budget before spending paid cloud credits.

• Implemented & Active Today:
  • Tailscale + SSH Mesh: Secure cross-network SSH staging between the Pop!_OS portable orchestration machine and the Windows RTX 3060 Ti workstation.
  • Software Factory Orchestration: Active multi-agent execution and experimentation using Oh My Pi (OMP) and T3 Code with isolated subagent workpools, alongside evaluation of Hermes Agent.
• Planned Features & Experimental Roadmap (In Progress):
  • Hybrid Cost-Aware Model Routing: Experimenting with intelligent task routing — delegating straightforward tasks to smaller, cheaper, or locally hosted models on the RTX 3060 Ti while escalating complex reasoning and architecture tasks to frontier cloud models across ChatGPT, Gemini, and GitHub Copilot subscriptions.
  • Persistent Linux Worker Nodes: Provisioning the two older HP Pavilion laptops as dedicated Linux nodes so autonomous background agents execute without interrupting active development environments.
  • Unified Cross-Device Orchestrator: Submitting high-level objectives from any device with automatic delegation across machines, models, and isolated worktrees, documented publicly on GitHub and X (@maseeek).

Loughborough University — BSc (Hons) Computer Science (2024 – Present) · First-Class Honours (Year 1 & Year 2)

Hills Road Sixth Form College, Cambridge — A-Levels: Mathematics (A), Physics (A), Computing (A) · GCSEs: 7× Grade 9, 2× Grade 8, 2× Grade 7

Languages & Systems      :: Python (Advanced) · C# / .NET · TypeScript · C/C++ · Kotlin · Java · SQL · PHP
Agentic & Orchestration  :: Oh My Pi (OMP) · T3 Code · Hermes Agent · CONTEXT.md / ADR Systems · Tailscale / SSH (The Lab)
Backend & Architecture   :: Microservices · PostgreSQL (RLS) · Node.js / Express 5 · Flask · REST APIs · OAuth 2.0 / JWT
AI, CV & Quantitative    :: OpenCV · Polynomial Trajectory Modeling · Anthropic Claude SDK · Gemini API · NumPy / Pandas · PyTorch
Frontend & Mobile        :: Blazor WebAssembly · React 19 · Next.js 16 (App Router) · Android SDK (Kotlin) · Tailwind CSS
Infrastructure & Testing :: Docker · GitHub Actions (CI/CD) · Vitest · Playwright · Linux (Pop!_OS) · Supabase · Vercel / Render