Missy
Production work
Most of what I ship at work is closed-source: production LLM agents, RAG pipelines, fine-tuned models and data infrastructure across fintech, edtech and retail. My resume has the names and dates; below are a few representative builds.
Agentic banking assistant—WhatsApp customer support built end to end: a LangGraph agent over RAG (Postgres/pgvector), with audio transcription, AWS OTP, scheduled LLM-generated financial reports, and an MCP server exposing internal tools.
Curriculum ML tooling—A transformer recommender surfacing semantically adjacent lessons with sub-second latency, plus a fine-tuned Llama 3.1 8B (LoRA) for resource generation that expert reviewers signed off as on-voice.
Retail data pipelines—High-volume AWS ETL (Glue, Athena, Lambda, S3) ingesting product-availability data from major retailer B2B platforms, feeding analytics that drove stakeholder decisions.
Open-source projects
More in my GitHub Profile.
Research & recognition
Understanding the Issues Surrounding COVID-19 Vaccine Roll Out via User Tweets—Peer-reviewed paper at CSoNet 2021, indexed in the WHO COVID-19 Research Database.
Empowering Julia-based Data Science with MLflow—Google Summer of Code 2023, bringing MLflow experiment tracking to the Julia ecosystem.
Convolutional Neural Networks in Respiratory Conditions Detection: A Systematic Review of the Last 4 Years—Research poster at SMILES 2020 (Skoltech), selected in the top 10% of 2,000+ applicants.
Notebooks
I rebuilt some classic Python notebooks in Julia using Pluto.jl for the reactive UX. Pick a topic.
MLOps & Lifecycle Management
High-Performance Computing & GPU Programming
Deep Learning Architectures
Data Engineering & Scalable Processing
Service
REPU Seminar 2026 Judge—Evaluated undergraduate research presentations for the Research Experience for Peruvian Undergraduates. April 2026.
IEEE Intercon 2024 Reviewer—Reviewed conference paper submissions for technical accuracy and originality. July 2024.