Selected work

Food recommendations, construction workflows, an offline ledger, and time-series ML.

Selected work

Full-stack · AI integration

FoodMind

Demo / staging

Food records, personal and group preferences, and recommendations brought together in a multi-client product.

My contributionHandled backend and web development, AI integration, infrastructure, and delivery throughout the team project.

  • Spring Boot
  • PostgreSQL
  • React
  • Python
  • AWS
One public boundaryOne public boundaryWeb + AndroidSpring BootPostgreSQLPrivate AI servicesRecommendationCookingChatBackend ownsPermissions · Rules · Validation

One public boundary

  1. Web + Android

    One shared public API

  2. Spring Boot

    Permissions · Rules · Validation

  • PostgreSQL

    Persistence

  • Private AI services

    Recommendation · Cooking · Chat

Conceptual system illustration

Enterprise workflows

ConstructIQ

Working prototype

A construction project workspace for tasks, progress reports, risks, team membership, and documents.

My contributionBuilt the Spring Boot backend and React management interface, including project access rules, Redis caching, and an AWS deployment workflow.

  • Java
  • Spring Boot
  • PostgreSQL
  • Redis
  • React
  • AWS ECS
A shared project contextA shared project contextProjectTasksReportsRisksDocumentsOwnership + membershipAccess and management are separate decisions

A shared project context

  1. Project

    Ownership + membership

  2. Tasks · Reports · Risks · Documents

    Access and management are separate decisions

Conceptual system illustration

Local-first · Data integrity

LedgerKit

Windows beta

An offline desktop ledger for multi-currency cash, investments, and net worth, with traceable changes and explicit data-quality checks.

My contributionDeveloped the React desktop interface and Rust application core, including decimal rules, import reconciliation, and encrypted backup and recovery.

  • Rust
  • Tauri
  • SQLite
  • React
  • TypeScript
One authority for financial rulesOne authority for financial rulesReact UIRust coreSQLitePreviewValidateCommitDecimal rulesTraceable changes

One authority for financial rules

  1. React UI

    Preview

  2. Rust core

    Validate · Decimal rules

  3. SQLite

    Commit · Traceable changes

Conceptual system illustration

ML · Anomaly detection

HGAD

ML experiment

Graph and hypergraph anomaly-detection experiments on server monitoring time series, from data preparation to model comparison.

My contributionImplemented data preparation, graph construction, autoencoder models, anomaly scoring, and batch experiments against several baselines.

  • Python
  • PyTorch
  • NumPy
  • scikit-learn
Relationships, then reconstructionRelationships, then reconstructionSignal windowsHypergraphAutoencoderAnomaly scoreExperimental pipeline · schematic only

Relationships, then reconstruction

  1. Signal windows

    Server Machine Dataset

  2. Hypergraph

    Build relationships

  3. Autoencoder

    Reconstruct signals

  4. Anomaly score

    Experimental pipeline · schematic only

Conceptual system illustration