Offline-first mobile app for disaster response. While every hackathon team built online, I identified SMS as the differentiator. Works when infrastructure fails.
Edge AITFLiteOffline-FirstMobile
Lance Jacob Silverio
Software Engineer. Fullstack & AI
Building fast and resilient systems from the frontend interface to local machine learning inference.
Skills
Core stack and tools I use to design, build, and ship production-ready software and AI products.
Languages
- Python
- TypeScript
- JavaScript
- SQL
- HTML
- CSS
Frameworks & Databases
- Next.js
- React
- FastAPI
- LangChain
- Supabase
- Prisma
- PostgreSQL
- SQLite
AI & Libraries
- LLM Integration
- Prompt Engineering
- NumPy
- Pandas
DevOps & Testing
- Git
- GitHub
- Jest
- Pytest
- Vercel
ML pipeline generating realistic synthetic PDFs. Multi-stage approach: deterministic data generation → LLM spatial reasoning → visual degradation. Solves synthetic training data problem.
ML InfrastructureLangChainLLMsData Pipeline
Calendar + announcements + members synced to Google Calendar. Shipped. Used by college orgs. Removes schedule fragmentation with two-way API integration.
Next.jsTypeScriptAPIsReal-time
Edge-first. Constraint-driven.
I'm Lance Jacob Silverio, a software engineer who builds for real-world constraints first. Edge AI pulled me in because it turns limitations into product decisions: low connectivity, low latency, and high reliability all at once. From offline disaster response to synthetic OCR pipelines, I focus on practical systems people can trust.
Principles
- • Design for failure before scale.
- • Use on-device intelligence when speed and privacy matter.
- • Keep systems structured so outputs stay reliable.
Building something with constraints?
Edge AI, offline-first systems, synthetic data, product strategy—happy to talk.