Jose Fernando Gonzales

AI Developer · Agentic Systems Engineer

I build agentic AI systems and LLM-powered applications — backed by deep FinTech expertise and a decade of turning business problems into production-ready technical solutions.

portfolio.py
from langgraph import StateGraph
from google import adk
def build_agent():
return
"employer": "JPMorganChase",
"stack": ["LangGraph", "Google ADK"],
"focus": "agentic_systems",
Python
LangGraph
Google ADK
JPMC
Enterprise AI in Production
Building agentic AI systems at JPMorganChase using LangChain, LangGraph, and Google ADK
10 yrs
FinTech & Systems Foundation
Deep experience in payment systems, solution design, and bridging business needs with technical implementations
Agentic
AI Architecture Focus
Multi-agent orchestration, RAG systems, and LLM-powered workflows — from design to production

Featured Projects

Personal and open-source projects that showcase end-to-end AI delivery. My professional agentic systems work at JPMorganChase lives behind the firewall.

Diabetic Readmission Prediction

Recall-optimised ML system to flag high-risk diabetic patients for discharge intervention — with SHAP explainability, fairness audit, and production API deployment

AIM x Emeritus capstone: end-to-end ML from raw clinical data to FastAPI + Streamlit deployment with LLM explanation chatbot

Python scikit-learn XGBoost SHAP FastAPI Streamlit LangChain Docker

RAG on Me (Resume Chatbot)

Intelligent resume assistant with RAG, guardrails, and source attribution for recruiters

Enables natural language querying of professional experience with accurate, contextual responses

Python LangChain Vector Databases RAG FastAPI

Quest to Solvin (RPG Chatbot)

AI-powered RPG chatbot with dynamic storytelling and context-aware responses

Showcases advanced prompt engineering and conversational AI design for engaging user experiences

Python OpenAI API Context Management NLP

Customer Churn Prediction API

Production-ready ML pipeline with RESTful API for business decision making

Demonstrates end-to-end ML deployment from business problem to actionable API service

Python Scikit-learn FastAPI Docker ML Pipeline

Gentleman's POS

Modular Python web application for dual-service business operations (carwash & restaurant)

Reduced transaction processing time by 60% across 2 businesses with real-world deployment

Python Flask SQLModel Docker ESC/POS Printers

How I Work

How I go from messy problem to working system

1

Understand

Define the real problem, success metrics, and constraints

2

Build

Develop. Focusing on practical business value

3

Ship

Deploy, measure impact, plan next steps

This systematic approach ensures every project delivers measurable business value while maintaining technical excellence.

See This Process in Action →

Latest Notes

Thoughts on AI, ML, and software engineering

All notes →
September 27, 2025

Quest to Solvin Case Study

Building an AI-driven interactive storytelling prototype with dynamic NPC generation and quest mechanics.

Read case study →
September 27, 2025

RAG-on-Me Case Study

A minimal RAG system that turns my resume into an interactive chatbot for recruiters.

Read case study →
September 27, 2025

Customer Churn Predictor API

End-to-end ML deployment: from data preprocessing to production API with Docker.

Read case study →

Open to AI Agent Engineer roles — let's talk

I'm open to select AI Agent Engineer roles where agentic systems, FinTech depth, and production AI experience are a strong fit.

Get in Touch