JOSEALVARADOALVARENGA

Senior Associate Software Engineer

Biotech → Software → AI

Scroll to see what I'm building↓

// SECTION 02

CREW BIO

Backend engineer with 3.5+ years building event-driven systems, data pipelines, and microservices in Python, TypeScript, and AWS at Capital One. Built a CLI that gives Claude and other AI agents on-demand access to centralized documentation across repos and services, and a Claude Code Skills marketplace to centralize AI tooling for engineers. On personal time, I'm building a Whoop MCP server that connects biometric data to Claude for personalized coaching and an AI workout app currently in TestFlight beta — including fine-tuning Llama 3.1 8B via LoRA for 95-99% cost reduction over the Claude API baseline. I studied biotech, which sounds like a left turn, but really I've just always been obsessed with how systems work and how to make them perform better. Now I'm training for an IronMan 70.3 to put it all to the test.

TECH STACK

Backend

Java

Python

PySpark

Microservices

REST APIs

AWS

DynamoDB

Lambda

Fargate

Glue

RDS / EC2

Multi-Region

Data

ETL Pipelines

Databricks

Event-Driven

Splunk

Mobile

Swift

SwiftUI

Claude API

MCP

Llama / LoRA

GenAI Tooling

Other

TypeScript

Node.js

Supabase

SQLite

OAuth 2.0

CERTIFICATIONS

AWS Certified Solutions Architect — Associate (2023)

// SECTION 03

SYSTEMS ONLINE

BETA

REPNOTES AI

PROBLEM

You're between sets with 60 seconds on the clock. Every workout app wants you to tap through menus and dropdowns to log what you just did. Logging a lift shouldn't take longer than doing one.

SOLUTION

One text input. Type "bench 225 3x5" and AI handles the rest, parsing exercise, weight, sets, and reps into structured data instantly.

KEY HIGHLIGHTS

◆Fine-tuned Llama 3.1 8B via LoRA for 95-99% cost reduction

◆Full ML pipeline: dataset construction, dedup, quality scoring

◆Natural language → structured exercise logs in < 1 second

◆11 beta users on TestFlight

STACK: Swift · SwiftUI · Claude API · Llama 3.1 8B · LoRA · Supabase

VIEW DEMO→TESTFLIGHT→

LIVE

WHOOP MCP SERVER

PROBLEM

You're training for a race with a Whoop on your wrist collecting HRV, sleep, recovery, and strain around the clock. All that data and the decision of whether to train hard or rest still comes down to a gut feeling.

SOLUTION

Connects your Whoop to Claude through MCP. Instead of staring at recovery scores and guessing, you ask Claude what to do today and it tells you based on your actual HRV, sleep, and strain data.

KEY HIGHLIGHTS

◆Acute-to-chronic workload ratio (ACWR)

◆HRV trend analysis

◆Cumulative sleep debt tracking

◆Race readiness scoring across 7 MCP tools

STACK: TypeScript · Node.js · Express · SQLite · OAuth 2.0 · MCP SDK

VIEW REPO→

INTERNAL

AI CONTEXT MANAGEMENT CLI

PROBLEM

AI coding agents are powerful but organizationally blind. They don't know your repo structure, your service boundaries, or where documentation lives. So they do the only thing they can, grep and read everything until something relevant shows up.

SOLUTION

A CLI for managing and serving organizational context to AI agents. Repos link in as submodules, creating a single source of truth across services. Agents and engineers get on-demand context at the org level instead of searching repo by repo.

KEY HIGHLIGHTS

◆Built and in active use across the team

◆Adopted by teammates, demoing across the org

◆Claude Code Skills marketplace for Travel engineering

◆Centralized AI tooling discovery and sharing

STACK: Python

INTERNAL — CODE NOT AVAILABLE

// SECTION 04

FLIGHT PATH

T-372DAYS TO RACE DAY

JAN 2026Training Begins

MAR 2026Building + Full Training← YOU ARE HERE

APR 2027IRONMAN 70.3 — RACE DAY← upcoming

CURRENT TELEMETRY

SWIM

1x / week

BUILDING BASE

BIKE

3x / week

ON TRACK

RUN

3x / week

ON TRACK

// SECTION 05

COMMS

EMAIL LINKEDIN GITHUB

Fueled by approximately 4,827 cups of coffee.