Skip to content
Paul Mandap

Open to software engineering roles

Paul Christian Mandap

Software engineer who builds things that can be checked — and then checks them.

I build full-stack applications, data pipelines, and local-first AI systems. What I care about most is the part after the thing works: proving what it actually does, finding the number that turns out to be wrong, and writing down the decision so the next person does not have to guess.

View workRésumé

01Selected work

View all ↗

Personal project · live, in use

Nomi

An installable study app, named after its companion, that turns a student's own notes into flashcards and quizzes that cite the page every answer came from, and grades written answers against a rubric. Nomi, an animated owl, chats, answers questions about your own progress without an AI call, and on one tap can build a study set from pasted notes or write a reviewer on a topic. Built for real users on free tiers.

952 tests passing
23.1k lines of TypeScript
13 tables, all RLS-guarded
TypeScriptReact Native WebExpo RouterSupabasePostgreSQL+6

Stratpoint Technologies · Data Engineering Internship

Movie Analytics Warehouse

An end-to-end ELT pipeline that turns three messy source files into a 20-model dimensional star schema, built for exactly one reader: a specialty-insurance underwriter deciding what premium to quote on a film.

20 dbt models · 3 layers
1m22s full warehouse rebuild
104 dbt tests
Apache Airflow 3.3dbt 1.12PySpark 3.5PostgreSQL 16Docker+2

Personal project

Personal AI OS

A model-agnostic personal agent system running entirely on a consumer GPU — no cloud provider anywhere in the codebase — with a purpose-built evaluation harness, a protected holdout split, and 69 architecture decision records including the ones that killed my own hypotheses.

1,058 tests · 1,042 offline
69 decision records
85 eval cases · 14 suites
Python 3.11OllamaSQLitepydantichttpx+2

Personal project · shipped to external testers

Nod

A Windows overlay that listens for a wake word, answers out loud in a locally-synthesised Philippine-English voice, and keeps a rolling summary of the meeting you're in — with microphone and speaker audio kept structurally incapable of reaching the same code path.

98.4% intent accuracy
606ms p50 classification
2–4s end to end, CPU only
Python 3.10PyQt6faster-whisperPiper TTSOllama+3

02Currently

Looking for

A software engineering role where the work has real users and the details have to be right — full-stack product work, backend, or data platforms. Manila-based or remote.

Building

A local-first agent platform on a consumer GPU, and the evaluation harness that keeps it honest. Most of what I learn comes from benchmarks that disagree with me.