Projects

Project Snapshot Run conversational AI experiments entirely on your laptop. This prototype couples a Hugging Face pipeline with GPU-aware guardrails so teams can iterate without depending on hosted endpoints. Business Context Targets labs and compliance-sensitive teams that need to validate chatbots without sending data to third-party clouds. Provides a repeatable launchpad for onboarding collaborators who work across Linux, macOS, and Windows. Core Capabilities Local BlenderBot inference driven by the facebook/blenderbot-400M-distill checkpoint via the transformers text-to-text pipeline. Notebook/script parity maintained with a Jupytext pair (basic_chat.ipynb ⇄ basic_chat.py) so edits stay synchronized across IDEs and browsers. GPU-friendly bootstrap that sets PYTORCH_CUDA_ALLOC_CONF and clears CUDA caches to squeeze models onto 2 GB cards while still offering CPU fallback. Environment diagnostics through gpu_ts.py and pt-cuda-ts, confirming PyTorch/CUDA availability before allocating large tensors. Conda-lock reproducibility with environment.yml, multi-platform lock files, and Makefile targets (make expenv, make updenv) that keep dependencies in sync. Implementation Notes Ships with concise setup steps for installing the right PyTorch build, transformers, accelerate, and sentencepiece. Encourages tight VRAM management by demonstrating how to toggle precision or device maps inside the pipeline. MIT-licensed and structured so larger checkpoints or UI wrappers (Gradio, Streamlit) can be swapped in later. My Role I packaged the notebook workflow, scripted the GPU health checks, and automated environment locking to make local LLM prototyping dependable for teammates. ...

Project Snapshot Sales leaders needed a trustworthy system to track incentives across cards, loans, and cross-selling campaigns. This backend streamlines goal management, reward policies, and compliance-ready reporting. Business Context Built to replace spreadsheets that obscured advisor performance and bonus eligibility. Equips managers with transparent dashboards to coach teams and justify payouts. Lays the groundwork for omnichannel banking experiences by exposing clean REST APIs. Core Capabilities JWT-secured authentication with Spring Security guarding every endpoint. Role-specific workflows for managers, advisors, finance analysts, and reporting stakeholders. Goal and points engine that ties products to reward tiers, updates progress, and prevents duplicate credit. Comprehensive API surface covering managers, advisors, financial products, clients, sales, and KPI reports. Realtime-ready architecture using WebSocket support for instant leaderboard refreshes. Implementation Notes Spring Boot 3.1 with JPA orchestrates persistence against MySQL, supported by H2 for local testing. OpenAPI/Swagger documentation auto-generated via SpringDoc for faster onboarding. Modular layering (controllers, services, repositories) keeps business rules testable and extensible. Packaged with Maven for streamlined deployment into bank-managed infrastructure. My Role I reimagined the legacy rewards process, modelled the domain, implemented the security and data layers, and configured observability scaffolding for future production rollout. ...

Project Snapshot High-volume data collectors often get blocked when Tor exit nodes appear on ban lists. This toolkit automates circuit hopping, preserving throughput without compromising anonymity. Business Context Ideal for research and compliance teams that rely on Tor to gather insights from restricted networks. Prevents stalled scraping jobs by detecting unresponsive or blacklisted circuits and rotating before requests fail. Core Capabilities Multi-circuit orchestration that spins up concurrent Tor connections for parallel requests. Circuit health monitoring to flag banned endpoints and trigger renewals instantly. Lightweight thumbnail generator guideline for showcasing Tor utilization in dashboards or portfolios. Designed as a personal library foundation that can be embedded into larger scraping pipelines. Implementation Notes Python-based helper utilities with clear extension points for queue managers or custom proxies. Emphasizes clean separation between Tor control signals and business logic to simplify auditing. MIT-licensed, enabling teams to extend safeguards or integrate alternative anonymity networks. My Role I distilled recurring automation needs into reusable primitives, focusing on resilient networking patterns and developer-friendly abstractions. ...

Project Snapshot Inventory operations for a fast-growing Peruvian beverage brand needed a single source of truth. I built a focused desktop logistics tool that keeps stock levels, movements, and alerts in sync across the company. Business Context Designed to replace fragmented spreadsheets with a guided workflow tailored to Oasisdrinks processes. Enables supervisors to visualize product turnover, pending restocks, and warehouse balances without manual reconciliation. Core Capabilities Modular Java Swing views for catalog management, purchase intake, and dispatch confirmation. MVC-based domain layer that isolates business rules from presentation for faster iteration. MySQL persistence with transactional safeguards to prevent duplicate or out-of-sync stock movements. Role-aware access that keeps critical actions limited to authorized clerks and managers. Implementation Notes Built with an extensible DAO pattern, paving the way for future web or API fronts. Includes seed data and environment scripts so new operators can trial the workflow quickly. MIT-licensed delivery with clean documentation for internal IT hand-off. My Role I led the full software lifecycle: gathering requirements with the operations team, designing the interaction model, developing the Swing UI and persistence layer, and deploying the pilot rollout. ...

Project Snapshot Buffet planners juggle recipes, budgets, and headcounts under tight deadlines. This application delivers a single workspace where chefs map menus, forecast costs, and adjust quantities without spreadsheet gymnastics. Business Context Designed for catering teams that need to simulate day-by-day menus for groups of any size. Bridges the gap between inventory oversight and client-ready proposals by surfacing real-time cost deltas. Core Capabilities Dynamic menu composer that calculates ingredient consumption per meal, per person, across multi-day events. Cost engine blending Java scheduling logic with Django services to present accurate purchasing budgets. User management microservice (Django) housed under login_micro, providing secure authentication and role-aware dashboards. Kubernetes-ready deployment, complete with Docker images and manifests for scalable rollouts. Implementation Notes Built with Java 17+ and Python 3.9, orchestrated via containers for predictable environments. Exposes REST endpoints that can power future mobile or kiosk extensions. MIT-licensed with clear documentation to accelerate onboarding for kitchen staff and finance partners. My Role I led solution architecture, implemented the Java planning engine, crafted the Django authentication service, and prepared the DevOps assets for container orchestration. ...

Project Snapshot Large-scale ticketing requires more than a monolith. This platform delivers fault-tolerant ticket sales through a suite of focused services, synchronized via NATS JetStream and packaged for Kubernetes deployments. Business Context Designed as a blueprint for marketplaces that need resilient shopping and payment flows. Demonstrates battle-tested event patterns, ready to extend into high-availability production environments. Enables rapid onboarding of frontend teams thanks to a modern Next.js client already wired to the APIs. Core Capabilities Microservice topology covering authentication, ticket inventory, orders, payments, and expirations. JetStream event bus embedded in the app for at-least-once delivery without extra infrastructure overhead. JWT-secured access with HTTP-only cookies and shared validation packages in common/. Responsive frontend built in Next.js + Tailwind CSS, delivering a premium checkout experience. Automated expirations that release tickets back to inventory when shoppers abandon carts. Architecture Highlights Dockerized services orchestrated via Kubernetes, with Skaffold scripts to streamline local dev loops. MongoDB stores transactional data while Redis can back transient caching or event replays. Shared TypeScript packages ensure consistent events, models, and error handling across services. Jest and Supertest guard critical workflows through unit and integration coverage. Implementation Notes Prioritizes observable, maintainable code: modular structure, dedicated namespaces, and defensive middleware. Real-time messaging flows documented for teammates stepping into the platform or extending integrations. MIT-licensed foundations that welcome customization for corporate environments or startups alike. My Role I architected the distributed system, implemented the service contracts, and hardened the DevOps tooling so teams can iterate quickly without compromising reliability. ...