Data & Machine Learning Analyst | Applied ML | Production AI Systems
I design and deploy end-to-end Machine Learning systems that turn real data into decisions — from data pipelines and model training to APIs and user-facing demos.
I come from a strong research background in generative models and signal/image analysis, and I currently focus on applied ML, analytics, and production-grade AI systems.
-
PRPD Synthetic Generator (Conditional GAN)
Generate synthetic PRPD patterns conditioned by discharge type.
👉 https://huggingface.co/spaces/jazg97/WGAN -
LLM-based Nutritional Assistant
USDA-grounded assistant with comparison, memory, and correction-aware dialogue.
👉 https://huggingface.co/spaces/jazg97/Nutritional-Assistant
- ZAP API — Internal Business Platform (Maintainer)
Maintainer of a production backend API supporting sales, inventory, cashback, commissions, and data workflows for a retail business.
👉 API Docs: https://zapi.avax.pe/docs
- Applied Machine Learning & Data Analysis
- Feature engineering, model validation, and evaluation
- Generative models (GANs) and Reinforcement Learning
- End-to-end ML systems: training → deployment → monitoring
- Data pipelines, APIs, and automation
| Project | Description | Demo / Status | Tech |
|---|---|---|---|
| PRPD Synthetic Generator (cGAN) | Class-conditional generation of PRPD patterns for data augmentation and robustness analysis | 🔗 HF Space | PyTorch, OpenCV |
| Traffic Signal Control (RL) | Reinforcement learning controller for traffic signals evaluated in SUMO | Replay demo in progress | PyTorch, PPO, Traci |
| LLM-based Nutritional Assistant | Grounded conversational assistant using public nutrition data | 🔗 HF Space | OpenAI API, Gradio |
| ZAP API (Production) | Internal business API for operations and analytics | Production system | FastAPI, PostgreSQL |
| Lung Ultrasound Classifier | Feature extraction and ML-based artifact detection | Offline research demo | PyTorch, scikit-image |
- Generative modeling (WGAN-GP, ACWGAN-SN, ACGAN)
- Reinforcement Learning for control and optimization
- Data analysis and feature engineering for real datasets
- LLM-based assistants integrated into business workflows
- Production APIs and data pipelines
I’m interested in roles such as:
Data Analyst | Machine Learning Analyst | Applied ML Engineer
- ✉ Email: *jose.zapana@pucp.edu.pe
- 🔗 LinkedIn: (coming soon)
- 🌐 Portfolio website (coming soon)