Skip to content
Oscar Navarro

Peru · Remote (UTC-5) — full overlap with US Eastern

Oscar Daniel Navarro Cieza

Frontend & AI-Augmented Full-Stack Developer

Four years building web and mobile applications in production. My core is frontend — React, Next.js and TypeScript — and I take responsibility for backend, database and deployment when a small team needs it.

View projectsDownload CV

Most of my recent work puts language models inside products people actually use. I deploy and maintain all of it myself, and I deliberately left several of these running unattended for weeks to find out whether they would hold up.

Based in Peru, UTC-5. Remote only, available as an independent contractor.

Projects

Screenshot of TickerLens

TickerLensAI equity research

Type a ticker, get a cited bull and bear case.

Every claim in the analysis links back to the source it came from. Retrieval runs over facts the application computes itself rather than over scraped text, with embeddings generated in-process and tokens streamed to the client as they are produced. Rate limiting is hardened against spoofed forwarded headers, and the app degrades gracefully when the database or the embedding model is unavailable.

Stack — TickerLens

  • Next.js
  • TypeScript
  • Tailwind
  • FastAPI
  • Python
  • PostgreSQL + pgvector
  • Docker
  • nginx
View live— TickerLensCode— TickerLens
Screenshot of Estudia

EstudiaAI study companion

Photograph handwritten class notes, get a study summary and an auto-graded exam.

A three-stage pipeline takes the photographs through transcription, summary and question generation. It uses context-window-aware chunking rather than retrieval, deliberately: the product needs full-document coverage, not a relevant slice. Multi-user with authentication, installable as a progressive web app and as an Android app. 168 tests, and a security audit whose ten findings were all fixed with regression tests covering each one.

Stack — Estudia

  • Python
  • FastAPI
  • SQLite
  • PWA
  • Android (TWA)
  • Docker
View live— EstudiaCode— Estudia
Screenshot of tvindicators

tvindicatorsstrategy validation engine

Trading strategies reimplemented in Python and run continuously in paper trading to build a live track record.

Over 40 validated strategies run against Binance perpetual futures, with a REST API exposing equity curves and live-versus-backtest evaluation. A pre-push hook blocks any commit where the live signal path stops reproducing the backtest. Paper trading only — it has never traded real money, and its own documentation warns against reading the backtest Sharpe as a live expectation.

Stack — tvindicators

  • Python
  • ccxt
  • SQLite
  • FastAPI
  • systemd
View live— tvindicatorsCode— tvindicators

Keplermarket-neutral trading system

A market-neutral crypto system that ran on real capital.

Seven uncorrelated alpha sleeves, leverage auto-sized to a maximum-drawdown budget, maker-only execution and a drawdown circuit breaker. Live maximum drawdown was 3.4% against a 10% budget across 18 days of live operation. Validation used walk-forward testing with purged out-of-sample data and embargo. Archived for business reasons rather than engineering ones: low-drawdown copy-trading at micro-capital produced no visible return without scale. It was a difficult decision, and I still consider it the correct one.

Stack — Kepler

  • Python
  • SQLite
  • FastAPI
  • Binance API
ArchivedCode— Kepler
Screenshot of Exposure Dashboard

Exposure DashboardOSINT aggregation

Give it a username, email or national ID and it reports public exposure with a confidence score and its reasoning.

Thirteen plugin-based collectors run in parallel with isolated failure handling, so one failing source never takes down a search. Identities are correlated across accounts by avatar perceptual hash, bio similarity and shared outbound links. Every redirect hop is validated against SSRF. 234 tests, running on a single vCPU.

Stack — Exposure Dashboard

  • Python
  • FastAPI
  • Next.js
  • SQLite
  • Playwright
  • Docker
View live— Exposure DashboardCode— Exposure Dashboard

Mi Pesolocal-first weight tracker

A weight tracker with no server and no account.

SQLite compiled to WebAssembly and persisted in the browser, with charts drawn by hand in SVG and no charting library. One codebase ships as a progressive web app and as an Android app. Its own SECURITY.md states the two honest limitations: the database is not encrypted on the device, and the published APK is signed with a debug key rather than a release key.

Stack — Mi Peso

  • React 19
  • TypeScript
  • Vite
  • SQLite/WebAssembly
  • Capacitor
View live— Mi PesoCode— Mi Peso

Also built: a desktop video editor for multi-gigabyte recordings with packet-accurate cutting and hardware-accelerated export, and a catalyst alerting service that measures the actual abnormal price reaction to each alert it sends.

Experience

  1. Frontend Engineer · Pairus Inc. (Canada, remote)

    Jul 2025 – Jan 2026

    Led the frontend of an AI dating product: personality questionnaire, embedding-based matching, real-time monitoring dashboards and Stripe payment interfaces.

  2. Software Engineer (contract) · Belsitec

    Jun 2023 – Present

    Admission and student portals for a university operating in Peru, Florida and Mexico; a full web platform and an attendance app for a private school.

  3. Frontend & Full-Stack Developer · Independent

    2023 – Present

    Reservation PWAs, self-manageable business sites with custom admin panels.

  4. Technology Consultant (internship) · IBM Peru

    Apr 2022 – Mar 2023

    Web and mobile development for corporate clients; IBM Cloud services.

Technologies

  • React
  • Next.js
  • TypeScript
  • JavaScript
  • Tailwind CSS
  • Python
  • FastAPI
  • Node.js
  • PostgreSQL
  • SQLite
  • Docker
  • nginx
  • Linux
  • Git
  • CI/CD
  • LLM integration
  • RAG
  • Progressive Web Apps
  • Capacitor
  • WebAssembly