⏰ What I’m Doing Now - October 2023

Here’s what I’m doing right now, inspired by Derek Siver’s /now page.

🗓️ Updated: 31st October 2023, Singapore

Currently writing this on Halloween so technically still on track for monthly updates!

🧠 Data Science

Still actively exploring permanent data science / AI opportunities in Singapore, a few interesting opportunities have potential - will see if they materialise to anything!

Had a coffee chat with the founder of a new AI consultancy, BrightRaven.ai, who welcomed me to the team as a freelance AI engineer consultant. They have some interesting LLM projects ongoing. Will see if I can juggle this on top of a full-time role.

Continue to write posts on AI and ML here on this blog, Substack, and Medium pages. I need to work on spending less time researching and writing, currently it’s taking me over a week to publish a post. Definitely guilty of Parkinson’s Law:

“Work expands so as to fill the time available for its completion”.

Working with my friend Son on a 30-day MLOps challenge building several ML solutions of varying complexity, here’s the first repo using MLflow on an AWS EC2 instance. Aim is to show data engineering and MLOps skills and tools to prospective employers, such as:

• working with APIs (requests, JSON)
• data processing (Spark)
• data orchestration (Airflow, Dagster, Prefect, Mage)
• stream processing (Kafka, Flink)
• cloud storage (AWS, Azure, GCP)
• experiment/model tracking and deployment (MLflow)
• CI/CD (Docker, Kubernetes, Kubeflow, GitHub Actions, Jenkins)
• making an impact through an API or dashboard visualisations (FastAPI, Flask, gradio, streamlit)
• good documentation/README
• other (unit testing, linting/styling, version control, linux)

Continue learning the latest AI advancements from meetups and conferences. I’ll be attending MLSG’s meetup on LLMs and RAG (Retrieval Augmented Generation) and also SWITCH startups event.

Technically not Data Science related but will help with blog writing and life in general - is that I’ve finally settled on a PKM (Personal Knowledge System) that works for me, consisting of Logseq (an open-source outliner alternative to Obsidian), Apple Notes, and Notion (for their databases feature for structured information). I recently came across the idea of a “Second Brain” by Tiago Forte and quite liked the idea of being able to easily access what you’ve read (books, academic papers, TowardsDataScience articles), thought, or even watched on YouTube - and seeing how all these ideas relate together through Logseq. Here’s how my graph looks after a few months of writing:

Logseq graph connection example by author

🏃‍♂️ Health

Will be travelling next month but after that I’ll be joining a climbing gym which also has great gym facilities too. Aim is to go at least 3x a week.

I’ve bailed on the last two running meetups due to the thunderstorms and lightning, fingers crossed it stays dry on the group running days!

📖 Reading

As well as finishing the two books from last month, I’ve finished reading Cal Newport’s Deep Work. Great book on producing deep and meaningful work and why it’s becoming more valuable today. Will take some ideas, such as the physical timeblocking, but not other ideas (some where a bit too extreme for me).

Reading Chip Huyen’s Designing Machine Learning Systems - to learn about the fundamentals of building resilient, scalable, maintable, and adaptable ML systems using MLOps best practices.

Currently 44% of my through The 4 Disciplines of of Execution by Chris McChesney, Sean Covey, and Jim Huling. Does provide a simple but hard to implement framework to achieving your “Wildly Important Goals”. Similar principals to OKRs from John Doerr’s Measure What Matters.

Started The Satsuma Complex by Bob Mortimer for some light-hearted bedtime reading.