Hey there, I’m Shanif. I’m a software engineer, data scientist, startup founder, and just someone trying to figure it all out. I write about it all here. Some of it will be about work. Some will be more personal. I can’t guarantee you’ll find any of it useful, but at some point there’s bound to be something good. If you haven’t done so yet, subscribe here.

I’ve started working on a lot of projects.

There’s Nobi, of course. I’ve started rebuilding Milton (a stock performance prediction product that was the foundation of the first project of my last startup, more on that soon). I’m building out an automated SEO engine that I’m hoping can form the backbone of an AI marketing platform down the line. I’ve also recently taken on a larger operational role with one of Nobi’s customers, helping them with AI automations and implementing better technologies into their business ops.

Any one of these could be full-time roles in and of themselves.

9 months ago my wife had a baby girl (super cute, when she’s not going crazy) — she wakes me up at least once a night. I’m also working with my family to try to continue to deliver on some of the social good/charitable work we’re doing (we have a couple of projects going with a few non-profit orgs, primarily in Pakistan, and have started to do more work to try to find opportunities to help).

So given all that, obviously, I thought the next thing I need to do is add a newsletter into the mix

Here’s the thing.

I’m learning a TON as I go, and some of the stuff I’m working on is pretty cool and can probably help others if they knew about it. I could also personally use a place to dump my thoughts. Writing is an outlet for me. When I get in the zone, I come out of it rejuvenated.

So I’m doing this. I’m calling it In Motion mostly because it’s pithy and I think it’s a good newsletter name, but also because it feels like I’m always pushing, grinding, trying to figure out what’s next. In Motion describes my life.

I’m not planning on writing on a set schedule like most of the real writers I subscribe to. I also don’t plan on writing only about one specific topic, nor do I plan on creating only valuable, educational, or generally helpful material.

Basically I’m trying to make this as tough on you, the reader, as possible 😉.

What I am planning on doing, though, is creating content that I think needs to be created. Oftentimes, that content will be helpful. It will be educational. It might show you how to do something that took me a long time to learn. Or it might entertain you or give you some life insights that you would have had to learn the hard way.

And if that sounds good to you, you should subscribe.

Let’s talk a bit about all this stuff I’ve got going on. It’s a bit crazy to be working on so much. You could rightly make the claim that if I’m not focusing on one thing, I’m not doing anything.

But the interesting thing about my life so far is that, more often than not, when I find myself working on a lot of things, something very lucrative/valuable/enjoyable almost always emerges from it.

That’s the belief that I have right now, and there is a reason I’m doing so much.

Back in 2023, my co-founder, Tyler, and I started working on a new startup.

LLMs were becoming popular and we both saw an opportunity to make something big. We tried to create and sell a product… it didn’t work. We sunset it and tried to go back to the drawing board. More interviews. More design partners. More selling before we built.

We eventually came across a repeating pattern that kept coming up in our interviews. Search and discovery was a real problem for retailers and ecommerce teams.

We built Nobi around that idea. It used concepts from RAG search to create a more effective and accurate search engine for retailers. We ran it in A/B tests on several ecommerce stores. It worked really well.

Turns out that nobody was chomping at the big to use it.

Some retailers wanted us to add merchandising capabilities for their website, outside of their search results. Others weren’t big fans of the chat-based interface, despite the fact that A/B tests showed it could drive more revenue. Yet others were locked into a contract with their existing search provider.

Search and discovery was a problem for them, but it was a latent problem that most folks couldn’t be bothered to solve right away.

Personally, I think they’ll change their minds. I’ve seen the numbers, our search is better than what they’ve got today. It also rolls in a support assistant that lets their visitors get help beyond simple product discovery. Retailers who implement it today get insights about what their customers are looking for, analytics around changing trends, and a great price.

I’m biased, I know. But it’s all objectively true!

But with our search & support product not taking off as quickly as we’d have hoped, Tyler and I knew we needed to find another problem to tackle.

AI is enabling so many new capabilities - we’re bullish on its capacity to change behaviors, and we suspect there are other opportunities out there for us to build something that lets us grow faster.

We still have a strong belief that the market will come around, and we’re seeing a lot of search demand and some inbound interest that gives us faith that this will eventually become a big product. So we’re pivoting to a world where we will hopefully capture some of the latent, inbound interest that we’re already seeing while we look for other opportunities to grow faster.

As part of making Nobi more of an inbound-led product, I worked with ChatGPT and Codex to figure out what kind of marketing plan we could use to draw interest (one which wouldn’t spend a lot of dollars, after all, if we had a bunch of money we could simply run the circuit at all of the ecommerce conferences).

The first thing it told me to do was update our website, which I did. I don’t feel great about it but at least it’s done.

The next thing it said was to really improve our bottom-of-funnel SEO (for those who haven’t gone as in-depth into SEO as I have, bottom-of-funnel is what essentially drives sales, there is a lot less keyword volume but a lot higher intent here).

But as I go forward, and I look to find tangential problems that might get customers to sign up for a solution more quickly than they do for search and discovery, I wanted to have a way to minimize any time I spend on Nobi, so I didn’t really want to commit to spending a ton of time writing articles every day.

Instead, as I looked at the overall marketing plan that I developed, and I knew that I might be experimenting with other products, I realized that there’s an opportunity to build out a fully autonomous, agentic marketing system for not only Nobi, but also every company.

So that’s what I started (and continue) to do. Back in April, I started working on an automated SEO engine, but one that would produce actually useful content, not AI slop. I knew that it would need to do a lot of things — keyword analysis, researching facts and competitors, outlining and creating articles, etc — so I started building out all phases of the system using Claude Code and Codex.

It’s now at a point where it has produced about 40 automated articles (see https://nobi.ai/blog for some examples). It’s not perfect yet, I still have to interact with it to get each article from 90% to 100%, but the amount of time that takes is significantly less than what it would take for me to write each article from scratch.

As they say, the last 10 (or 20) percent of progress in most projects requires 80% of the effort, so that’s where I am now with this SEO engine. There’s a lot of tweaking and tuning to do to get it to a point where it’s fully automated, I’m not sure it’ll ever really get there, but at least where it is now is already adding a ton of value.

As part of discovering other areas where there might be potential for fast growth, I’ve started to revisit a project I started about 10 years ago - stock forecasting using machine learning and article analysis.

For those of you who knew me back then, this was the start of the Milton product (which was the concept behind Apteo’s first effort). The idea actually worked really well - we took hundreds of millions of data points about stocks, including textual articles, and we used them to forecast future stock performance. Our predictions resulted in early detection of some huge winners (think NVDA in 2019), but we shut down the product because we couldn’t figure out how to get people to use it (one of my biggest learnings in startup life so far is that distribution is the startup).

In any case, now that coding agents have become pretty good, I’m having Codex rebuild Milton in the background, and also improving some of the things I wish we had done the first time around. I’m hoping to get the first constructed portfolios out in the next couple of weeks.

I’m thinking of hosting everything at miltonnau.com (“nau” being the transliteration of the word that means “new” in some Middle Eastern and other languages). I have’t done it yet but I’ll probably get around to it once the data science and backtesting platforms are in place. I might also open source it, but I’m still thinking through that.

On top of all that, I’ve started working/consulting heavily with one of Nobi’s first customers, doing more to help them implement AI in their operations. Actually, one of the more exciting things about this is that I can start to implement some new features for Nobi that I’ve been pushing back on, because they might benefit from those features, and I can start to take the automated SEO engine I’m working on and expand it to apply to them.

I must say, I do like building things. Now, more than that, I like building things that people actually use and making lots of money to do so, but at the very least, I find myself working on things that are very interesting for me, and I am doing my best to keep that going.

Throughout the past few years I have paid myself primarily through crypto and options trading (maybe more of that later). I’ve found some ways to make it work for me, and for that I’m very grateful.

I’m still trying to figure this all out - hopefully finding something that I can turn into a highly profitable business, trying hard to be there for my new daughter, trying to maintain the few (but extremely tight) relationships with important people in my life now that I’m in my 40s.

The only thing I know for sure is that I’m figuring it out, but I think… so is everyone else, and for that, we’re all in motion and enjoying this journey.

More to come when I have more to say.