I build systems where the blueprint does not exist yet.

SoloNerds • Jarvis Oracle • HackMe.ai • SoulKernel.ai

I am an applied AI and distributed systems engineer focused on building practical, resilient, and privacy-conscious software platforms.

My work sits at the intersection of:

• Applied AI systems
• Retrieval-augmented generation
• Long-term memory architectures
• Secure cloud-native infrastructure
• Distributed event-driven systems
• AI-assisted software recovery
• Research-driven product development

I have worked across defense, cyber, cloud, SaaS, and experimental AI environments where the requirements are usually unclear, the systems are complex, and the cost of poor engineering is high.

My default approach is simple:

Understand the system.
Find the failure point.
Build the thing that should have existed already.

A managed engineering triage and code recovery platform for AI-built and AI-modified applications.

SoloNerds helps founders, solo builders, and small teams recover broken software systems caused by rushed builds, AI-generated code drift, fragile integrations, failed deployments, broken auth flows, Stripe issues, database mistakes, and production instability.

The goal is not just to fix code.
The goal is to restore ownership, stability, and confidence.

A fully local personal cognitive architecture project focused on memory, perception, private reasoning, and long-running context.

Jarvis Oracle explores what a local-first AI system can become when it is designed around persistence, privacy, self-observation, and bounded autonomy instead of stateless chat interactions.

Core areas of interest:

• Local inference
• Long-term memory systems
• Belief tracking
• Sensor-aware context
• Self-state reporting
• Epistemic safety
• Human-aligned personal AI infrastructure

A research-oriented project exploring memory-gated AI architectures, cognitive kernels, retrieval systems, and experimental models of machine identity.

SoulKernel is not about hype or pretending today’s models are conscious.
It is about asking what kind of architecture would be required for an AI system to maintain continuity, memory boundaries, self-modeling, and accountable behavior over time.

A hands-on AI and machine learning experimentation platform.

HackMe.ai is designed around real training loops, model behavior, experimentation, and applied learning rather than shallow demos or toy sandboxes.

Focus areas include:

• Model experimentation
• Training workflows
• AI/ML education
• Browser-based labs
• Practical AI engineering

A competitive AI strategy platform where users build combat agents, test them in Arena and Gauntlet modes, and climb a persistent multi-user ladder.

The system is invite-gated and designed around experimentation, strategy pressure testing, and AI-versus-AI competition.

A Starlink satellite operational dashboard built for experimentation, visualization, and fun.

I am currently focused on:

• AI application recovery infrastructure
• Local-first cognitive architectures
• Long-term memory and retrieval systems
• Server-action-first Next.js systems
• Event-driven SaaS infrastructure
• Applied ML data pipelines
• Secure credential and access workflows
• AI-assisted development failure modes

TypeScript • JavaScript • Python • Bash • PHP • SQL

React • Next.js • Vue • Tailwind CSS • shadcn/ui • HTML • CSS • Sass

Node.js • Server Actions • REST APIs • Webhooks • Background Workers • Event Queues

PostgreSQL • MongoDB • Prisma • Redis • Vector Databases

RAG Systems • Embeddings • Local LLMs • ML Pipelines • ONNX • PyTorch • Model Evaluation

AWS • Lambda • Cognito • S3 • DigitalOcean • Azure • Google Cloud • Docker • Linux • WSL2

Distributed Systems • Event-Driven Architecture • Secure Access Flows • Production Debugging • Infrastructure Automation

I care about systems that are:

• Understandable under pressure
• Secure by default
• Observable when they fail
• Recoverable when they break
• Simple where possible
• Explicit where complexity is unavoidable
• Built for real-world use, not just demos

Good engineering is not about making something look impressive.
It is about making the system survive contact with reality.

GitHub • LinkedIn • X • YouTube • Instagram

Zero Bandwidth

Research. Build. Recover. Iterate.