$
whoami
Ali Fardaev
▋
$
cat
about.md
Ali Fardaev — ML/AI engineer focused on taking models from research to production.
He builds end-to-end systems: data pipelines, model training, serving infrastructure, and monitoring.
Outside of work, he enjoys exploring nature and photography.
$
cat
education.md
UC Berkeley — Master of Information and Data Science (MIDS)
Focus: ML at Scale · ML System Design · NLP · Data Engineering · Deep Learning
$
ls
projects/
MedCoverage
Python
Helps Medi-Cal patients check whether a medical procedure is covered —
converts dense policy documents into a short yes/no questionnaire and
returns a clear covered or not-covered verdict with policy reasoning.
RAGFAISS
FastAPIRedis
LanceDBAWS
Flight Delay Prediction
Python
Predicts flight departure delays at scale — 30M+ OTP records joined with NOAA
weather data, engineered with temporal-safe features and trained with Spark MLlib
MLP on a 48-core Databricks cluster.
PySparkSpark MLlib
DatabricksDelta Lake
MLP
Vizomaly
Python
YOLO fine-tuned on custom anomaly data — detects visual defects in images
with a dataset curated and labeled specifically for anomaly detection.
YOLOFine-tuning
Anomaly DetectionComputer Vision
Ignis AI — Wildfire Risk
Python
UC Berkeley MIDS capstone — predicts wildfire risk from satellite imagery
using deep learning, enabling early detection before fires spread.
PyTorchSatellite Imagery
Deep LearningGeospatial
$
ls la
links/
$
echo
$CONTACT/