ENGINEER · BUILDER · OPERATIONS PROBLEM SOLVER

AI Solutions Engineer & Full-Stack Developer

I build AI-powered software, intelligent workflow automation, and full-stack applications that eliminate repetitive work, improve operational visibility, and help businesses make smarter decisions.

WORK ORDER ROUTE
  1. Plan
  2. Schedule
  3. Execute
  4. Improve
ASSET PLANP-1204AStrategy · Reliability
MARKET INTELLIGENCE
⌁╱╲╱╲╱
Signals · Context · Risk
PRODUCTION APPWorkflowsAPIs · Data · UX
Maintenance planning + SAP
Market intelligence
</> Production applications

SELECTED WORK

The problem comes first.
The technology follows.

Five systems that show how I move from a human or business need to architecture, analysis, interface, deployment, and measurable usefulness.

Maintenance Planner desktop application monthly planning screen
01

DESKTOP APPLICATION · OPERATIONS

Maintenance Planner

A planning system built from firsthand manufacturing experience to turn SAP work orders into balanced, technician-ready weekly schedules.

Why I built it

Maintenance planning contains a large amount of repeat work: exporting SAP data, reviewing priorities, checking technician availability, accounting for production blocks, balancing hours, and rebuilding the same weekly views. The application moves those repeatable decisions into software. It applies consistent scheduling rules, produces CSV output for SharePoint visibility, and gives the planner more time for exceptions, coordination, and higher-value reliability work. The goal is not automation for its own sake—it is less administrative effort, fewer avoidable planning errors, and a clearer plan for the people executing the work.

  • SAP connection and Excel import workflow
  • Trade, shift, skill, and capacity-aware scheduling
  • Production blocks and rotating-team calendar logic
  • Plan export for execution and handoff
C#.NET / WPFSAPScheduling LogicExcel
Olmem Market Intelligence dashboard showing market conditions and stock opportunities
02

FULL-STACK APPLICATION · MACHINE LEARNING

Olmem Market Intelligence

A market research platform that combines live market context, stock scoring, portfolio-aware insights, news signals, IPO tracking, and optional AI explanations.

Why I built it

Stock research is fragmented across price screens, company news, market indexes, sector performance, portfolio positions, and new IPO information. That makes it difficult to understand what deserves attention and why. Market Intelligence brings those signals into one decision-support experience. It helps a person compare opportunity, momentum, pullback quality, and risk; monitor established companies and upcoming IPOs; and optionally ask an AI provider to explain the evidence. It does not place trades—the purpose is to make research more organized, timely, and understandable before a person makes a financial decision.

  • Rule-based scanner with momentum, value, pullback, and risk scores
  • Python machine-learning layer for model refinement
  • Bring-your-own AI provider integrations
  • Authentication, subscriptions, alerts, and portfolio tools
Next.jsTypeScriptPythonPostgreSQLAI APIs
Crypto trading scanner showing evaluated assets and guarded buy decisions
03

TRADING SYSTEM · AUTOMATION

Autonomous Crypto Trader

A self-hosted trading application that scans liquid crypto assets, evaluates momentum and risk, and can execute guarded Coinbase trades around the clock.

Why I built it

Crypto moves continuously, and a person cannot responsibly watch every asset, trend change, spread, and risk condition twenty-four hours a day. The trader was built to monitor that volatile environment consistently and respond to evidence rather than fatigue or emotion. It compares current momentum with historical behavior, volume, moving-average direction, liquidity, spreads, and account-level risk limits. Explainable blockers show why a trade is rejected, while trailing exits and loss controls protect the process. The objective is disciplined, repeatable decision-making in an always-on market—not a promise of guaranteed profit.

  • Live market scanning with explainable buy blockers
  • Risk limits, trailing exits, spread, volume, and loss controls
  • Manual override and automated trading modes
  • Persistent server-side settings and trade reconciliation
Node.jsJavaScriptCoinbase APIPostgreSQLDocker
spotify_ensemble.py
# Select interpretable cohorts
silhouette_results = []
for k in range(2, 11):
  model = KMeans(
    n_clusters=k,
    random_state=42
  )
  labels = model.fit_predict(X_scaled)
  score = silhouette_score(
    X_scaled, labels
  )

# Combine diverse learners
ensemble = VotingClassifier(
  estimators=[
    ("rf", base_models["Random Forest"]),
    ("et", base_models["Extra Trees"]),
    ("gb", base_models["Gradient Boosting"]),
    ("ada", base_models["AdaBoost"])
  ],
  voting="soft"
)
BEST ROC-AUC0.898Extra Trees
SOFT VOTE ACCURACY86.85%4-model ensemble
04

PYTHON · ENSEMBLE LEARNING · DATA SCIENCE

Song Cohorts & Marketing Ensemble

A repeatable Spotify marketing analysis that groups songs by listening characteristics, then ranks marketing candidates with supervised ensemble models.

Why I built it

Listeners rarely choose music from technical audio features, but their behavior reveals patterns in the energy, danceability, acoustic character, mood, tempo, and style they prefer. The project turns those patterns into useful cohorts so a music provider can suggest related songs, shape playlists, and test marketing against likely listener interests. Unsupervised clustering answers who a song may appeal to by grouping similar tracks. Supervised ensemble learning answers which songs deserve marketing attention by estimating their relationship to higher popularity. Together, the two layers support more relevant discovery instead of recommending music based only on a single popularity number.

  • Cleaned and validated 1,610 tracks, then engineered interpretable audio and release features
  • Standardized 15 clustering features and used PCA to visualize the feature space
  • Selected two K-Means cohorts by comparing silhouette scores across k=2 through k=10
  • Benchmarked Random Forest, Extra Trees, Gradient Boosting, AdaBoost, and a soft-voting ensemble
  • Selected Extra Trees at 0.898 ROC-AUC; the soft-voting model reached 86.85% accuracy
PythonpandasNumPyscikit-learnMatplotlibSeabornJupyter
Why these libraries
pandas + NumPyCleaning, feature engineering, aggregation, and efficient numerical work.
scikit-learnReusable preprocessing pipelines, PCA, K-Means, ensemble models, and comparable evaluation metrics.
Matplotlib + SeabornEDA, correlation analysis, cohort interpretation, and model-result communication.
olmemtech.com
OLMEM TECHNICAL SOLUTIONSProfessional websites.
Custom AI systems.
Website solutionsAI systems
05

BUSINESS PLATFORM · PRODUCT ENGINEERING

Olmem Tech

The customer-facing platform for my software business, presenting website, CMS, workflow, and AI automation services with product intake and account experiences.

Why I built it

Small businesses often live with disconnected spreadsheets, repetitive communication, hard-to-update websites, and manual processes because generic software does not match how their work actually happens. Olmem Tech exists to begin with the business problem, understand the current workflow, and build the smallest useful system that removes friction. The platform demonstrates that approach while giving customers clear paths to website solutions, custom applications, CMS products, and AI automation based on the need they are trying to solve.

  • Built and shipped the complete responsive website
  • Productized custom software and AI service offerings
  • Customer accounts, intake flows, and payment integrations
  • SEO-focused service architecture and conversion pathways
Next.jsTypeScriptTailwindStripeVercel

CURRENTLY BUILDING

New solutions for work that should be easier.

Active product directions shaped by recurring business problems—not technology looking for a use case.

01

AI Business Automation

Turns repetitive research, intake, follow-up, and internal handoffs into controlled AI-assisted workflows.

02

Market Watch AI

Makes business, stock, news, portfolio, and IPO information easier to compare and understand.

03

Manufacturing Workflow Software

Reduces repeated planning, scheduling, reporting, and coordination work around production operations.

04

Intelligent Chatbots

Helps customers find accurate answers, qualify needs, and reach the correct next step without replacing human judgment.

05

Custom Enterprise Applications

Replaces disconnected spreadsheets and manual processes with software shaped around the actual operation.

THE DIFFERENCE

I understand the need before I choose the technology.

I came to software engineering through industrial maintenance, automation, and planning—where the best solution begins by listening to the people doing the work and understanding what is consuming their time.

Today I combine that operational judgment with full-stack development, data science, and AI. I do not begin with a preferred framework or an AI feature. I begin with the repeated task, fragmented information, missed opportunity, or difficult decision; find the constraint that actually matters; and build something people can use.

Manufacturing OperationsMaintenance PlanningSAP ExperienceWorkflow OptimizationAI EngineeringFull-Stack DevelopmentMachine Learning
10+ yearstechnical operations
0 → productionend-to-end ownership
Business + codeone problem-solving mindset

EXPERIENCE & CAPABILITIES

Technical depth grounded in operations.

2024—NOW

Maintenance Planner

SC Johnson · Components Manufacturing

Plan and coordinate preventive and corrective work across multi-craft teams, production constraints, contractors, and regulated manufacturing operations.
2019—2024

Electromechanical Technician

SC Johnson

Troubleshot, maintained, and improved automated manufacturing and packaging equipment in high-uptime environments.
Product engineeringNext.js, React, TypeScript, JavaScript, Python
AI & dataGenerative AI, agentic workflows, PyTorch, pandas, scikit-learn
SystemsREST APIs, PostgreSQL, authentication, payments, SAP integration
DeliveryGitHub, Vercel, Docker, DigitalOcean, testing and production support

LET'S BUILD SOMETHING USEFUL

Looking for an engineer who can connect business problems to working software?

I'm pursuing AI solutions and full-stack software engineering roles where ownership, curiosity, and practical judgment matter.