Building a AI Agent to analyze REITs and InvITs

 Ages ago, I came across the news that the minimum threshold of 100 units for REITs to buy have been done away with. I was curious what is REIT? Why the limit size and so on. There were only 3 REITs then. 

I was fascinated by them. I felt it might be interesting cause it helps to diversify, it helps to own a piece of real estate and regular income and yields were better than rental yields. (Think buying a flat and renting out) In addition, potential for capital appreciation as well in a clean and better manner. (Think about buying a plot and the loan and the hope it will appreciate and the challenge of liquidating and so on)

Though the lot size limit was done away with, I bought 100 units of all 3 REITs. I kept buying them regularly. I wasn’t sure of figuring out which is good and which might have potential and so on. So I bought all the three regularly and more or less in equal amounts. Now and then I would try to search for posts and analysis about them but they were never up to date.

Fast forward now, thanks to AI, generative AI and LLM, I have been building out few things. (How I ended up as a vibe coder is a story for another day) One of the things, I wanted to build was to figure out the retail bond with the best yield. I was struggling and end up facing structural challenges. It was then I realized, I could use that approach to analyze the REITS.

Basically, I wanted to know which REIT could be the best one to pick rather than blindly investing equally in all REITs. 

https://rankindiareitinvit.lovable.app/

So I have built it. Also the dividend yield is only one factor and mostly the past which may or may not help to look in to the future. So, I wanted it more of a strategy engine, what If i am keen on future potential, what If i am looking to reduce the risk, what If i have to Optimize for the dividend yield (again based on the past performance)

During the process of building it I learnt more than few things.

One, just because prompting and tools make it easier to build and gives a nice looking output doesn’t mean it does what it is intended to do. We should also know what we are building, why we are building and the nuances of it. Domain knowledge, Software engineering, product management matters more than ever. 

For e.g as simple as fetching the CMP was a challenge. In the first cut, it was fetching stale data from yahoo. Calculating Yield is one thing, calculating it the post tax yield is what makes it unique. There are nuances to tax calculation of REITs, in addition all REITs are not unique. In fact, though I was thinking it is better than rental yield, never thought about, hey what is the bench mark to see, if the yield is good (again to be fair, it is not just about yield as there could also be capital appreciation). 

Even when building it, I asked to fetch the parameters from the investor presentation and AI was taking it from the PDF. Considering, I am someone who knows about the industry, I should have asked the AI agent to fetch the data from the XBRL rather than the PDF as it is the optimal and ideal one.

Once I was done with building, the surprising thing was how the REITs were not necessarily doing well as the post tax yield was below the 10Y G-Sec Yield. 

Later on I also added the InvITs to the mix. (Again the taxation for InvIT is different) 

I could have also built this differently. For e.g have only one page with filter to have REIT only, InvIT only or both together. Rather than the current version of 5 pages, - Master ranking, REITs, InvIT, terminology and methodology, and what this is all about. 

I also learnt the following from Lovable: Feedback on how I prompt

• Skip explaining the problem — just state the desired outcome. I can see the code.
• One instruction per sentence. Your multi-option prompts ("do X, or alternatively Y, I prefer Y") cost extra tokens exploring options you don't want.
• Reference by component name when possible ("TerminologyCard: use inline badges") rather than describing the UI.
• Batch related changes in one message with numbered bullets — your audit prompt was a good example of this.

Roughly, your prompts averaged ~2-3x the minimum needed length. Tighter prompts = fewer tokens per message in build mode.

What next?

I could make it open source, I could build a admin panel or one with user input, where people give the inputs as country, taxation, local nuance, basically say where to look for the APIs to fetch the CMP, NAV and the returns, and it builds the engine for that market. 

I can try and build a plugin to calculate the taxation and make it easy to file the taxes in India.

Few other things that might be of interest

I had around 18-22 high intensity prompts. Probably would have spent 120K-150K tokens and have spent 8-10 hours. I could have achieved the same in 30% fewer prompts. If upfront, I had thought a lot about how to build, what to build, say i want both REIT and InvIT, I want the table to look this way, these are the common factors, this is how I want to visualize and so on. 

Why Being Around Smart People Isn’t Enough

 Nowadays, I intentionally try my level best not to use my mobile phone. Digital de-addiction and all that.

It also means, lot of time, I am introspecting about past, future. Sometimes, I also worry and may be a little bit scared that Doom scrolling is better than these kind of thinking :) Just Kidding!

The reflections led me down this rabbit hole!

what is my weakness? what is the thing that is so stupid about me?

"I have access to really smart people and am not acting upon their recommendations!!!"

Long Long ago, I once hosted someone from Valley for a bunch of folks. Someone who has written an academic book on Machine learning which is recommended text book in Stanford and all. Post the meetup, I felt the conversation and discussion was sub par and I wasted his time.  A year later, he reached out and asked can I set up a similar meet up. I was puzzled and asked him, why is he keen. He said, "Last time we spoke about X and Y. I was not at all aware of them, though they are right in my backyard. I made a note, I should reach out to them but totally missed it and now as you know they are such a big deal. I really feel that I missed out and probably this meet up again would help me to unearth some big bang opportunity. 

Too many takeaways to me like, how the valley folks could live in a bubble, how even valley folks miss out on the trends and so on. Most important was the pattern matching and how he thought about how to look for opportunities.

Now coming back to my life

I have a mentor, who has advised me on my career choices, whenever I am at the cross roads. Each and every time, I have taken a detour from his recommendation, only to reach at the destination he recommended. If only, I had listen to him, it would have saved me a lot of time and considering the detours, they all would have compounded well. 

I have another friend, who is really darn good at looking in to the future, in to the edge/curve and all that. Sometimes, I think I am not sure, even he knows how good he is at that. 

Circa mid 2016-17 time frame, when all the VCs were chasing the shared rides, quick commerce and all that, he was asking about blockchain startups and how they are looking at it and so on. I should have bought bit coin and held on to them. It was below $500

Fast forward the post-covid crash, I was holding a stock, which had fallen more than 50% in value. He was strongly recommending me to sell it and I would still be better off the stock and would come on top and all that. I didn't listen to it. Now it has fallen further 50% and today it is less than 1/5 of what it was. Even if I had sold the entire portfolio then and invested in a index, I would have made more than for all the loss incurred earlier

2 years back, around 2023, he said, "AI will disrupt software engineering in ways we could never imagine" He said that in 2023, before 2 years. He was managing a sizable team and delivering AI assisted/generated code at enterprise level for mission critical customer facing applications. He said, development time and roadmaps will get compressed by factor of years with half the team sizes. How much I wish I had started learning AI and had a head start in the AI learning. 

So,

One, go and find such good and smart folks in your life, if you don't have them

Two, if you already have them, count your blessings and listen to them!

What Would Jobs Do? - A Custom Steve GPT!

 TLDR:

I have built a custom GPT: What Would Jobs Do?  

https://chatgpt.com/g/g-68f7a59494c481918995e7eecb050617-what-would-jobs-do-wwjd

Becoming a Steve Jobs Fan

When I was 11 or 12, a neighborhood senior kid told me about Apple, Mac and Steve Jobs. I genuinely thought he was bullshitting me thinking that I don’t know about Bill Gates. 

Fast forward, when I wanted to buy a gift for someone very close during my first US trip, a friend was like, if you really want to gift properly and on a budget, get them a iPod mini. It was the entry product and still it was too expensive when compared to other mp3 players. I was like, what this is on budget? 

I was curious and went down the rabbit hole and that is how my fascination and adoration with Steve Jobs got started. It was too strong, I ended up writing a book about him in Tamil. It is been more than 10 years and now and then I still get fan mail for it.

The unfinished (actually unstarted) book on Steve Jobs!

The unintended consequence of deep diving about someone like Steve Job is, me losing my mind over when people (read tech bros, founders) channel their inner “A” and passing it as modeling the Jobsian behavior. Each and every time, I read an article referencing Jobs and the bad behavior of founders, I was like, guys, come on.!

So for a long time, I have always wanted to write a follow up book on Jobs in English. Well actually two. One, was a compilation of all his quotes. Two was to write about how one could learn from Jobs to make a dent in the universe rather than being a painful “A” I wanted to highlight the lessons, takeaways, and how he had evolved as a person and how his best was at his later point of time after learning his lessons.

I had collected, read, made notes on all the books that were available, created a playlist of his interviews running for hours.

I started compiling all his videos, interviews, books that I missed earlier and the books published after my book. (I had to rely a lot on iCon and lot of early articles across publications like Forbes, BusinessWeek, Time, Wired and Fortune to write my book; The Walter Isaacson wasn’t published when I was done with mine)

However, I never got around to writing them. I couldn’t get around to getting the transcripts from the YouTube video complilations. 

Eureka! Eureka! How about a SteveGPT?

With the advent of chatGPT, and NotebookLM, I thought I could give this a shot again. I almost succeeded. ChatGPT helped me to get the transcript of the videos. NotebookLM made the researching from the books, the compilations, pretty easier. They compressed the time, it would have taken me to manually go through my notes, compiling, by a huge factor.

In the meanwhile, I also had to become AI literate and was reading a lot and seeing a lot. I also realized, the best way is to get the hands dirty, and actually build something so I can truly understand how it works. (It is one thing to lead vs actually build)

It was then it also occured to me, I could write a book, get to lot of people, yet people may struggle to apply or make it relevant for them. As one drama coach told me, “sometimes, an event/act is like a dot. There are 360 degree POV to see them”

I started wondering, what If I just don’t use AI, GenAI and LLMs to write a book but also potentially to actually help folks. For e.g people who write self help books, productivity books, run a community, where the readers/audience share their problem, learning’s, and people customize their learning’s. Well, the authors also leverage the readers mail as their content. Most of Dilbert’s strips were inspired by what the readers wrote to Scott Adam’s.

What If I could compile, consolidate, all of these things to model Steve Jobs behavior.

Building the SteveGPT

Initially, I was naive to think, I could just upload the books, give the links of the videos, and my old collection of material and voila the GPT would become a smart Steve Jobs.

Then came the learning’s on how the GPT works, need to chunk, how it stores, how it generates, how to build the knowledge base and at the end of the day, the output is a function of input, and how it really works and contributes to the mission of making the Jobs behavior, advice and input, personalized and customized.

Hopefully, unlike the book on Steve Jobs, sometime soon, I will write in detail on how I built this custom GPT and learning’s and how one could build their custom GPTs.

The name and the irony of the GPT

In the macintosh days, the team always wondered, How would Steve Jobs react to it? The team created WWJD. What Would Jobs Do? More like a play on the famous, “What Would Jesus Do? More like a dig at his messiah like complex then.

However, in the later years, when he handed over the rein to Tim Cook, one piece of advice he shared with him was something to the effect of, "never ask what I would do" and instead to "just do what's right". He often quoted that Disney fell in to trap because people always thought, what would walt do!

Anyways, it is okay to ignore the Jobs advice, and whenever you are in Jam, whenever you think you need the counsel of Steve Jobs like persona, ask this GPT, “What Would Jobs Do?”

https://chatgpt.com/g/g-68f7a59494c481918995e7eecb050617-what-would-jobs-do-wwjd

Build, Buy or Vibe Code…

 I had just joined the corporate incubator program. I was full on enthu mode to prove myself, make a mark, do this, do that, change this, change that. Super supportive boss definitely added more fuel to my enthusiasm.

Among the many other things that I was raring to do, two things were of high priority to me.

One, to rebrand the program. Rebranding was promptly shut down (However, it eventually happened, and it is a story for another day)

Two to modernize and transform our application process. Startups had to download the word document, fill and mail back the application. As a tech company and tech mentorship as one of the core offering, we all were aligned, we could and should do better.

Boss was supportive, offered budget and all the resources and the empowerment to do the right thing and get it done. Promptly get a tech team to build the application portal and whole shebang.

With all the buzz and discussion going on, one of the first time founder with a rich experience, asked me what I am up to. Full of pride, I told him, we are building the application portal. 

He went on to offer his two cents and said, why not buy some of the offerings available readily in the market. He further went on to say, you also have to maintain it, and need to have someone for it and it will add to your overhead in all aspects. You are better off focusing on aspects to help us rather than spend your time on this. (I get the irony that, ideally I should advice the founders and not the other way around)

The builder in me  was like, how can a tech company as big as us with all the talent around, buy something rather than build. 

Eventually, relearnt the lesson, one should never build if it is not your core and you are better off by buying it. So, when we expanded, to a new geography, we just bought an application management system.

However, I would repeat the same mistake of building another offering for internal use case (again for a good reason) and eventually buy something from the market. (Again story for some other time)

I can’t help but wonder, with all the vibe coding tools available, what would I do today?

P.S: I am a proud vibe coder :) 

Google Keep and Google Gemini

Few days back, wrote a post on good to have things in the NotebookLM. One of the call out was to integrate Keep and NotebookLM.

Actually Google does it one better.

Google Keep and Gemini integration is seamless and super cool. 

Once you enable, Google Workspace under the connected apps in the settings of Gemini, the Keep and Gemini are integrated.

Post that, you can query about your Google Keep from Gemini.

Say for e.g you have multiple to-do lists. I have a label todo for work and personal and maintain a list based on days. I can ask Gemini to give me a consolidated deduplicated task list.

In fact, you can even ask Gemini to add items to your to do list in specific label and it will do that as well. 

who are you and where are you?

 I wrote about the power of peers and that led to me to further thinking...

There is more nuance than to just saying peers. It is about the ecosystem, the environment and all that it encompasses.

Think about the following:

Can Cheetah win a swimming competition?

Can a fish survive in land?

Can crops grow in seasons and environments that are not conducive to them?

Can a seed grow on a rock or when the conditions are not favorable?

So when you succeed, it is not just your effort, your capabilities, your smartness, it is also about lot of factors and the environment you are in that helps you to execute, that helps you to deliver, that nurtures you

Similarly, when you are not firing on all your cylinders, probably, may be it is not about you. Probably, you are a polar bear suffering in the sub-saharan african desert.

Just make sure you are in the right place!! 

Power of peer groups

 One of my friend would often quote the Roger Bannister Phenomenon. Basically until Roger broke the 4 min mile record, people thought it was an unachievable feat. Once he broke, it was like floodgates opened and people started running the mile under less than 4 minutes.

You basically need someone else to show that can be done and everyone thinks it is doable.

It plays out more than ever we could ever imagine or observe.

Think about the following

In certain ecosytems, you would do see a lot of people doing the similar or same thing. Going to the same foreign country for job prospects. Schools sending the kids to a particular college, people from the same college starting successful firms and so on.

As they say, you are an average of 5 people whom you spend time with. 

Though I have heard and have seen this behavior (For e.g I have seen people preparing for Masters in US or applying for specific companies or doing MBA cause their peers were doing) Yet, I was pleasantly surprised to realize I haven't internalized it as much as I should have and was genuinely surprised to see it in action.

For e.g I meet my writing mentor occasionally like once in a blue moon and every time I see him, he would be like you would not continue to write as you are not moving in a circles and as a result you would not be motivated to write. No drive no inspiration and all that, I used to think, he is little upset I am not in touch with him.

However, in recent times, (well not so recent) I started playing wordle. Well, I started playing cause most of people I knew were playing it. Once I started playing other games in NY times app,  others too started playing and it was becoming like a friendly competition. The most interesting thing was, when i played strands, i would use hints and when I saw others not using and solving it, I also started doing the same. I was terrible in connections (I am still) but when i started seeing others solving at ease, even my solve rate has increased before that. 

It is just games in an app yet there is a powerful lesson in human behavior. 

If we see others doing things, we think and believe we too could do the same. 

So it all depends where you are, what you see and what you end up imbibing. 

If you are in a high achievers group, you too are motivated the same way. The irony is, let us say if you are not in high achiever group, the odds of you falling down to the group level than you raising the level of the group is very less.

beware of your peer group. beware of whom you spend your time with and interact with. Talent density matters.

As the cliched saying goes, if you are the smartest in the room, then you are in the wrong room...

Selling AI...

 If you are someone selling AI or AI enabled/powered solutions, it would be good to bear the following in mind.

• Most end customers or end users think AI is all powerful. Magic wand and can solve all the problems. They would say, "This is the problem and it can be solved in this way using AI and why is that it is not solving"  They may not realize that their systems may not be good enough or capable to leverage their AI or their data not good enough to train or be more reliable or even there are better and simpler alternatives than AI
• Most client firms could also possibly have their own tech teams and they might be competing you with intensely than ever before. 
• Most of the clients firms probably have burnt their fingers by the half baked promises and unfulfilled promises and as a result would look at anything you say with lots of suspicion
• It is better to call out what are the features you have already built and what is on the roadmap. Again clients would want to know what is vaporware and what are empty promises and how much of that is real
• Case studies, validations, references and testimonials could help, but you are served by adding caveats as you would not have a better and actual insights on the state of data and how they are organized and how easy or difficult would the training be and how quickly you can get to the 100% reliability. In other words, be transparent, and if anything undersell rather than oversell

As the saying goes, more the things change, more the things remain the same...

What I want from Google...

 Am definitely excited by the Gemini 3 launch as much as any other tech bro...

However, all I would want from Google is the following enhancements to the NotebookLM (as they say in Tamil, that that man that that problem)

1. Ability to upload doc and epub formats. I have many ebooks in epub format and it would really be a great help to upload them directly to NotebookLM

2. Sync up with Google Keep directly. Google Keep is simple, powerful. It is my brain dump, digital note taker. How much I would love to sync my Google Keep directly with NotebookLM

Doing the right thing

 Was reminded off these two events. 

Things being cryptic are intentional!!

Bunch of us were evaluating a bunch of firms. 5 firms to be precise. Most of the firms were represented by the stakeholders themselves. One of the firm had made an exception and sent their senior person. 

None of us were bothered that an agent was sent on behalf of the principal. 

Across the board with all the 5 firms, the discussions were casual. More open, more informal. More friendly. As a result, guard was down and people were speaking freely.

This exceptional firm who had sent a senior person, made an off handed remark, in fact offered an information unsolicited, something to the effect of "Oh we don't do X because it is risky" Basically the person made that statement to showcase they are robust, they have best practices. 

The room was full of men. One of them, responded, "Thank goodness, She (mentioned the name) is not here, if not you would have had a tough time" The aforementioned "She" was the final decision maker for our evaluation. 

This exceptional firm was really exceptional, they were the largest, best, would have ranked the top in all metrics we were evaluating. 

At the end of the day, we huddled up internally for the review to prepare the recommendations before presenting to the decision maker.

The analyst who had prepped the deck outlining pros and cons for each firm had just mentioned, "Not culturally aligned" across this exceptional firm. None, literally none said anything about how exceptional this exceptional firm was or how they were actually crushing. All of us moved on without spending another second. 

Another Incident

Bunch of us were at a dinner. It was me who had organized that dinner with internal and external stakeholders to get to know each other and could help us with one of the critical initiative at the time.

One of the external stakeholder was unnecessarily rude to one of the waiters. It was odd, yet could be considered par for the course. Not always, folks treat the support people in a nice way and if they think not being nice is okay, they behave that way. 

One of my senior stakeholder pulled me aside and said, ensure we don't engage with this person. 

Good culture and Good values and Good principles are about doing the right thing, when no one is watching and even when stakes are really high!!!