AL
About
Iterating on AI systems, amplifying organizations and teams.
I design, build, and operate ML/recommender/LLM systems on the job.
In my free time, I tinker on personal prototypes below.
Portfolio
llm-evaluator
nextjs
typescript
fastapi
python
postgres
speech-to-text
text-to-speech
llm
Obsidian Copilot
github.comeugeneyan/obsidian-copilot
Copilot for writing & reflecting; based on hybrid RAG.
typescript
python
docker
opensearch
fastapi
retrieval
UX for RecSys + LLM
eugeneyan.comwriting/llm-ux/
Prototype UX integrating an LLM into a recommendation system.
python
docker
fastapi
recsys
llm
Visual Search
eugeneyan.comwriting/image-search-is-now-live/
Image search for fashion, furniture, etc; deployed at Lazada/Alibaba.
python
theano
flask
transfer learning
Testing ML
github.comeugeneyan/testing-ml
How to test ML code on implementation, behavior, and performance.
python
docker
pytest
testing
evals
typescript
gatsby
interviews
guides
papers
Industry ML Meetup
meetup.comml-meetups-virtual/
Monthly meetups with ML practitioners & paper authors.
meetups
hosting
papers
facilitating
Specialties
Tabular ML
Drives 80% of value for most businesses.
Recommendation Systems
Improving discovery, conversion, revenue.
Language Modeling
Personalizing text content for users.
ML Engineering
The 80% of code that's not ML code.
Organization Effectiveness
Amplifying teams that build systems.
(Technical) Writing
Gaining clarity, scaling ideas.
Experience
Amazonmlrecsysllme-commerce
2020 - Now
Senior Applied Scientist
UCare.aitabular mlsequential modelinghealthtech
2018 - 2019
Machine Learning Lead
Alibaba/Lazadarecsystabular mle-commerce
2015 - 2018
VP, Machine Learning
IBMtabular mlrecsysconsulting
2013 - 2015
Data Scientist
• Anti-money laundering, social media analytics, recommender systems
Stack
python
scala
typescript
nextjs
docker
aws
spark
pytorch
opensearch