Hello and Welcome!

Who I am, why I write, and why you might be interested

First of all — thank you so much for reading this post! While I’ve been writing on Medium for a couple of years now, I haven’t yet gotten to introducing myself on Medium. I’m a data scientist but my journey has not been typical. I started off getting my PhD in physics, doing research, working my way to be a professor in a small college, and writing data science articles here on Medium and mentoring folks along the way. I think writing on Medium in top publications like Towards Data Science, Towards AI, Start It Up, etc. enabled me to smoothly transition to the data science industry. Over the past years, I’ve learnt a lot on Medium — what content works well, what doesn’t, and how to be successful while having fun writing!

Oh and I’m also a dad to twin boys and love to spend quality time with them as well! Believe it or not — my interactions with my sons help me in writing blogs, communicating knowledge, and building larger audiences. This summer, we are taking them swimming a bunch — and are seeing the fruits of their/our hard labor. A couple of weeks back, they were scared of getting in the water — but because of dedicated efforts from their end, as well as some positive feedback from our end — they are finally comfortable with their pool noodles at the deep end! But the ride has not been completely smooth. We have endured our fair share of screaming — before because they were scared, now because they don’t want to get out of the pool!

Consistency adds up, and efforts pay off in the long run. Even if that doesn’t seem true in the moment. Sometimes the best you can do is to stay positive.

What Do I Write About?

Before diving into how I made it here, let me talk about what I write about and why you might be interested. Over the last year, I’ve become proficient in applying state-of-the-art natural language processing (NLP) and AI methods to industry use cases. Some of my more popular articles include topics related to transformers, open-source vs closed-source large language models, and also geospatial data science:

If you are interested in how to apply the latest and greatest AI advancements to industry use cases, you are in the right place. I spend a lot of time researching these latest advancements from all sorts of sources — Research papers and company announcements, social media, courses, etc. I then experiment with these tools — and write accessible blogs for a wide range of folks. Non-technical folks can skim these articles, and technical folks can read through in detail. I also provide code snippets and GitHub links for those of you who are interested in replicating the results/applying to your own use cases. I’ve had many folks reach out on LinkedIn regarding questions on how to use what I demonstrated on their use cases. So feel free to reach out!

My Journey Through Blogs

As promised — I’ll talk about how I reached here, chronicled through my blogs!

The first blog I wrote was in late November, 2019 — about research I was doing as a post-doc, modeling the impacts of large-scale cyber attacks on vehicles. The blog had a very small reach, but I got an email from The Startup — requesting whether I would like to publish my blog in their publication.

At the time I had no idea about the blog — but I realized it was the biggest Medium publication at the time!

That single experience made me seriously consider writing regular blogs. After this, I published my first blog in Towards Data Science about traffic patterns:

Next, I wrote a series of blogs on the “science of everyday phenomena” — which also corresponded to a course that I was teaching at the time and an open-source book I made for the course. I created a new publication “Emergent Phenomena” for these articles:

While these articles on everyday science were not super widely read — they provided me with a routine and consistency by which I paced writing and publishing. It was also during COVID — where the college shutdown for quite a few months and I spent a lot of time writing.

I was getting ~2–3k monthly views at this point, and mentoring beginners in data science — when I decided to transition to the industry. I wrote a blog about this as well:

Soon after, I started experimenting with NLP and fine-tuning transformer based models for custom applications. So I wrote a blog on this:

After this blog, my monthly views became more like 8–10k. Since then, I have written multiple blog articles on NLP, AI, and industrial applications:

When Should You Fine-Tune LLMs?

LLM Economics: ChatGPT vs Open-Source

How Do You Build A ChatGPT-Powered App?

Extractive vs Generative Q&A — Which is better for your business?

Fine-Tune Transformer Models For Question Answering On Custom Data

Unleashing the Power of Generative AI For Your Customers

Build Industry-Specific LLMs Using Retrieval Augmented Generation

After a few such articles on Generative AI to solve real-world business problems, my monthly views rocketed to 30k+!

The Future — A Roadmap

I plan to write exclusively on AI and industry applications over the next year. I’ve seen that this is a niche that is very important because of all the folks trying to adopt Generative AI (ChatGPT/open-source LLMs) into their business needs. There are great folks making fundamental innovations at OpenAI, Google, etc. and writing research papers as well as communicating this research to the public. At the same time, folks at places like Hugging Face and AWS write brilliant tutorials on deploying state of the art open-source LLMs on their platform.

However, there is a lack of information on how to integrate these fundamental innovations to specific industry settings. For example, one thing that was really lacking was a detailed price comparison between open-source and closed-source LLMs. Or a discussion about what sort of LLM to use for what case. This is where I plan to focus my content (and have done so over the last ~10 or so blogs).

So stay tuned as I explore the nitty gritties of AI, Deploying models at scale, cost vs quality, and lots more — packaged in a way that is tailored to solve business needs. I also created a YouTube channel a while ago (where I’m now going to be more active on).

In the next phase, I’m going to create a landing page once I have a better idea of curated content I can provide (e.g. courses, a weekly newsletter on the latest and greatest AI advancements relevant for businesses,…) — I haven’t yet nailed this part down yet, but comments are really appreciated!

I look forward to having you all join in my journey of learning, creating and sharing knowledge!

If you like this post, follow me — I write on topics related to applying state-of-the-art NLP in real-world applications and, more generally, on the intersections between data and society.

Feel free to connect with me on LinkedIn!

If you are not yet a Medium member and want to support writers like me, feel free to sign-up through my referral link: https://skanda-vivek.medium.com/membership