Why do you promote RDBMs?

(Originally posted on my linkedin on Sep 8, 2017)

Following is based on a true conversation…

Smarty Pants (SP) : Akki, which is better? RDBMS or No-SQL?

Me : Which solution are you building?

SP : This is an enterprise solution

Me :  You have business data?

SP : Yes

Me :  Business transactions?

SP : Yes

Me : Are these transactions part of a business process? With people/teams and systems come and go at different stages?

SP : Yes

Me :  Guess, you need an RDBMS

SP : Why do you say RDBMS is better than No-SQL?

Me :  I said you need RDBMS for your problem

SP : Why are you promoting RDBMS?

Me :  I am not promoting RDBMS but that is what is suitable for your situation

SP : Explain

Me :  You will need strong ACID support. And so, support for normalized db store

SP : Thats cliche and beaten to death. You are old and prejudice.

Me :  Really? Which part? Your solution doesn’t need ACID support or normalization is not required?

SP : Why do you assume you can’t do these in a NoSQL?

Me :  But why would I even look at NoSQL? Thats meant for different set of problems. How do you ensure transaction boundaries and something like 2-phase commit in a NoSQL?

SP : See, you do not know enough about NoSQL but have decided so quickly. Which NoSQL db you are familiar with?

Me :  I worked with a couple. Mongo is what I am more familiar with.

SP : Which version?

Me :  Don’t remember. Downloaded 6 months back. Whichever was latest. Needed a store for unstructured data and Mongo is really good

SP : You don’t even know which version of db you are using. Mongo 3.x version has something called isolation that can give you transaction boundary and guaranteed commits.

Me :  Good to know. But I guess that is supported for special cases. If you do that with all business transactions in the system, I would expect it to be very expensive operation in not normalized, redundant, huge documents of MongoDb. Because MongoDB is not built for that.

SP : Again you are being prejudice. Why can’t you normalize in MongoDb? You can have your data normalized and have isolation to ensure your transactions.

Me :  What?? if I am going to normalize my data in a NoSQL and mimic RDBMS transactions, why should I even use MongoDB? Its an apology for NoSQL/MongoDb. They are created to solve totally different problems

SP : You are not moving with times. You are not updated.

Me :  Really! Why do we have an Excel and Word? You can have all graphics, text, formatting etc., within Excel itself? Why? Am I favoring one to other if I pick Word for text document and Excel for a spreadsheet?

SP : You are so adamant. Many like Facebook are using noSQL. You know that?

Me :  I know and they are doing that and rightly so. They are not building an enterprise business solution the way you described your problem. I will be curious to see where their internal financials are managed.

SP : You are stuck in 90’s and not appreciating new technologies

Me :  I give up. And glad I am not born after 1985! And haven’t learned computer science from wikipedia.

-Akki (No plans to retire any time soon)

Posted in Big Data, NOSQL, RDBMS, Startup | Leave a comment

AI Path from ANI to AGI may follow Predator-Prey Model

(Originally posted on my linkedin on  Aug 12, 2017)

Recently I read this 2-part write-up “The AI Revolution: The Road to Superintelligence”. I know enough of AI to understand this elaborate writeup and fat claims within. However, I fall short to be able to assess the validity of those claims. But that doesn’t stop me from ramble around.

I feel that this write-up has missed discussing the currency of ‘emotion’ that humans use so eloquently in their transactions. Which has to be tamed (IMO) by machines to go from ANI to AGI. Indeed, there is a brief mention of this aspect and to quote,

‘To be human-level intelligent, a computer would have to understand things like the difference between subtle facial expressions, the distinction between being pleased, relieved, content, satisfied, and glad, and why Braveheart was great but The Patriot was terrible.’

But the above is underplayed on how this might slow down increasingly as ANI (Artificial Narrow Intelligence) converges to AGI (Artificial General Intelligence).  Let’s look at this using a recently published Facebook research on making chatbots negotiate with each other.

Facebook research explains us how two machines ‘logically’ arrive at most optimal result at both ends.  This is machine to machine negotiation.  If it is a human to human negotiation, we employ many other items like emotional blackmailing, sentiments and manipulation. Favoritism, racism, exploitation etc., are all realities of life. Is it rare we say ‘man we both are from same school, you will make that extra for me’ or ‘you look like a sensible lady and you will understand my situation, you will give me for the price’.  And people may fall for these. The optimal solution that machines would arrive will have no relevance or computational explanation in such situation.

If a machine has to negotiate with a human, all these traits must be mastered by the machine to negotiate (manipulate?) him/her. Not impossible, it only takes providing sizable such transactions and transcripts between humans as a learning data.

But the challenge begins, as we start taking out machines, first at one end of the table and then at the other end.  At some point there is no more human left for machines to learn from human behavior. Once humans are out of the equation, the evolution of these machines may fall short to appeal to humans. That may prompt humans to come back into the fray.  And machines will learn again.  So, as ANI starts to spread across various fields, it’s likely that it will learn more from fellow machines and less from humans.

First generation of driverless cars will evolve based on other human drivers on the street. As the number of driverless cars increase on the road, they learn from each other’s evolution and behavior.  Imagine street design, construction etc., are also done by machines. Machines that drive cars and machines that build infrastructure learn from each other. There is no human needed here.  It should be intuitive by now that at some point these machines may go far away from observing, factorizing and/or prioritizing how humans are reacting to these decisions.  There is no bad intention within these machines, they learn from the training data that is available to them and humans are not contributing to that dataset anymore!

This may result in a situation similar to  ‘predator-prey model’ .  Humans become the prey and machines become the predator. That is within the context of the model and not literally (hopefully).     As ANI goes far away from humans, humans will jump in and take over. As more humans start interacting, ANI becomes smarter again and reduce human interaction. This cycle will continue but how long?   I guess that ANI will become AGI the day the wavelengths of this predator-prey model converge to zero. What I am not able to conceive in my thought experiment is, does the wavelengths increase or decrease as the time passes by?  If the wave lengths have to increase, ANI may never become AGI!

Links

  1. https://waitbutwhy.com/2015/01/artificial-intelligence-revolution-1.html
  2. https://waitbutwhy.com/2015/01/artificial-intelligence-revolution-2.html
  3. https://code.facebook.com/posts/1686672014972296/deal-or-no-deal-training-ai-bots-to-negotiate/
  4. https://en.wikipedia.org/wiki/Lotka%E2%80%93Volterra_equations
Posted in Uncategorized | Leave a comment

Innovation is the purpose, not a cultural element!

(Originally posted on my linked on Aug 5, 2017 )

Yesterday attended a brilliant session ‘Building leadership, organization culture and capabilities for innovation’ at #PLF2017HYD . Thank you all the good folks at Product Leadership Forum for a worthy day!

Wonder if “innovation” is a characteristic of an organization’s culture. Don’t know why we started to say this. Guess, it is around 2007/2008 that the whole of Indian IT industry woke up and decided to build this “organizational culture to innovate”.    By this date 2007/2008,  Indian IT & ITES market size was 70+ billion USD. In 2016 it grew up to 150+ billion USD.*

An industry that is making 70+ billion dollars and whose primary work force is ‘engineers’ accumulated all this worth without innovation?  Something is not adding up right!

I think innovation is never an issue. And it is never part of any “culture”. Innovation is THE PURPOSE. That has been and has to be the reason why IT industry exists and thrives.

Agreed, there is a shift around 2007/2008 and industry had to pivot.  But please, that is not to ‘start innovating’.   Change was in deciding to innovate ‘what’ instead of ‘how’.  Until then the focus was on processes to solve someone else’s problem. Change was to identify and solve problems on its own.  That was the shift.

Looking back, probably even I am guilty of thinking and phrasing Innovation being part of an organization culture.  At some point I personally was driving one such initiative to train entire engineering department to identify ideas and experiment. That is training the workforce to do their job. I don’t want to call it a culture shift.

We got to make innovation as the purpose and reason why we get up in the morning and go to work.   Leave dress code, open/closed door policies, flat/hierarchical departments,  alcohol/no-alcohol etc., to organizational culture.

* Source : https://www.ibef.org/industry/information-technology-india.aspx

Posted in Uncategorized | Leave a comment

Don’t say YES to Big Data if you want to say NO !!

Deck I used in key note at Great India Developer Summit 2015, Hyderabad

 

Posted in Big Data, Systems Of Engagement, Systems Of Record | Leave a comment

Lessons for Marketplace startups from Uber Episode

First, it pains to read about the unfortunate incident involving Uber in New Delhi.  As a citizen of this country, I hope and demand the law of the land puts its best foot forward to bring justice to the victim and culprits to the book. 

The unfortunate incident also raises some important questions about emulating successful business ideas from west in India.  It takes little effort to find a list of   ‘market place’ startups trying to organize the unorganized sector in India.   Be it providing domestic services or household goods.  

Relatively,  it costs little and technically easier to bring up a site to match services and consumers.  The challenge however is if the business has to endorse the behaviors of vendor and the consumer, one to the other.  That is beyond the need to have a sight on the quality of the service/product being transacted.

Imagine if the business has to send a plumber to a household.  The business has to endorse not just the quality of the service but also the behavior of the plumber at the household.

Similarly, if the business has to send a female maid to a household, business has to endorse the behavior of the customer towards the service provider.  

Verifying a service provider could be relatively easy but not cheap. 

  • The back ground checks needed
  • Strong lawyer to formulate disclaimers and contracts to share the risks
  • Operational model that defines strict processes and policies.  And adhering to those policies and procedures.

All the above are  going to cost a fortune for a business.   The startups typically do not think about or understand the costs involving these. 

Verifying the consumer may never be fully possible because its not practical to have a background check done on an entire population. 

These are some of the lessons we learned from our failed startup Makemydabba.com.    We enabled  home made food delivered to  needy end customers at work and home.  In no time we realized that we are enduring a sizeable risk of making unknown people meet at places we have no control on through our portal.   This is beyond the quality risk of the food being served.

We thought about models to mitigate those risks, but all of them cost a lot more money involving lawyers, our own delivery models and more.   That’s where the big funds needed to keep the business afloat.  A lot of people continue to ask us why a successful startup like Makemydabba had to close down for lack of funds.  Above is probably the primary reason why more funds were needed.

Coming back to the incident involving Uber, they cannot claim to be a startup.  They must have learned their lessons at the places where they are successful.  How they failed to bring them into India is something we should wait and see as the investigation process takes its course.

In addition however, we must acknowledge that India is relatively a low trust society.  We cannot trust a certificate, license, permit to be genuine on its own merit.  Even the most trivial of the information like an ‘address’ cannot be trusted without a ‘proof’ or someone certifying it via a background check. 

Even Indian regulations prefer to be stay gray for a business to know if it is doing right or wrong.  For example, we ran from pillar to post to understand if  Makemydabba  requires a food license or not.   We still do not know the answer.  We consulted numerous lawyers, CAs,  experienced food business houses and of course the kind mentors in Hyderabad.  No government office / official will say with authority how you should register your company and how you can be fully legal.  But all of them will come after you the day something goes wrong.

Its extremely important for startups in this space to take this incident as an eye opener and be cognizant  about the risks involved.   Simply put, if your business is enabling two unkown people to meet, the business has to put in place systems to take the risk completely out.   That is, risk to consumer, vendor and you the business! 

Posted in Uncategorized | 3 Comments

Specific tools that can be readily used.

(This is the review I wrote a while back on Flipkart)

 
This book from Vinay Dhabolkar & Rishikesha T. Krishnan “8 steps to Innovation” belongs to the category of books that gives us specific tools that can be readily used. It is in the league of books like Switch and The Lean Startup. It is written in the Indian context but the principles can be applied elsewhere too.

In this book, we get to find specific tools like creating a challenge book, democratizing innovation, designing and running rapid low cost experiments, iterating the business model and many more. There is a danger in writing such a book. A heavy weight process manual may claim the same “tools for you to succeed”. They may also claim that the tools prescribed by them are an amalgamation of all the “best practices” discovered elsewhere. The problem with best practices is that they fail in narrating the context. This is where the new breed of books like the “8 Steps to Innovation” differs. Each tool that gets introduced in this book comes along with a detailed narration of a true story thus laying out a strong foundation for context. Thus the tool becomes a “bright spot” (a concept mentioned in the book Switch) and not a best practice.

Being a first-time entrepreneur (during my YAssume/MakeMyDabba days), there were two kinds of books that attracted my attention.

In the first category are the books that inspired and motivated me. These inspired me to take up the challenge of creating something new. Books like Founders At Work, Stay Hungry Stay Foolish, Black Swan, Outliers fall in this category. Some of these are biographical and anecdotal. Others are conceptual and theoretical.. Such books play an important role in shaping one’s intent and intellect to take up a bigger/newer/different challenge. These books provided me the encouragement to jump out of a corporate job and to become an entrepreneur. However, once you hit the ground, not many of these books come in handy to deal with real-life situations. There is a danger that one loses hope quickly and starts to conclude that he or she is not a maverick enough like the celebrated ones figured in these books.

There comes for the rescue, the second category of books. That’s where I list this book “8 steps to innovation”, prominently alongside books like Lean Startup and Switch. At MakeMyDabba (second of my startups), we managed to take the idea of creating a web platform to help people at home to list and sell home food to others in their neighborhood, from drawing board to a successful launch in less than 8 weeks. We owe a lot to books like “8 Steps to Innovation” that provide hands-on tools to perform rapid experimentation and to get to customers faster. Such books can easily and proudly claim to be a part of the startup culture!

Posted in Uncategorized | Leave a comment

Applying Systematic Innovation to Customer Development

Here is the slide deck I used during my talk at AgileTour 2013 Hyderabad.

Here is the abstract of the session

Lean Startup model teaches us how to make a beginning in  conditions  of extreme uncertainty.   The concept of  Minimum Viable Product (MVP) and Leap Of Faith Assumptions are devised to keep the validated learning as the main objective.   What follows is the Customer Development  which is all about identifying the customer and collecting the right feedback to course correct. In real world situations it is common that the feedback collected can be overwhelming and  daunting.   Systematic Innovation is a set of tools that can be used to prioritize and pick the right feedback to work on.   Systematic Innovation also helps us on how to generate ideas, how to design experiments and more.  This 45 minutes lecture quickly introduces how to marry these two Lean Startup and Systematic Innovation.

-Akki

Nov 30, 2013

Posted in Uncategorized | Leave a comment

How you fund your startup in Indian Context (Super Idiot Guide)

(I thought I will keep quite and stay away from startup/entrepreneur world after shutting down my ventures.  As it turned out, I became more active and vocal.  Evidently, more and more youngsters want to try a “startup” these days.  Some of them are approaching me to find answers on how to go about starting the new journey.   I am trying my best to share whatever I managed to learn in the past 3 years.)

Disclaimer : I have no qualification in this subject let alone being an expert.  Following is a very high level view from someone who started as a novice 3 yrs back and forced to understand some of this.  Target audience for this post are those who come from pure technology back ground with an idea of trying a startup but do not know how to go about funding the adventure.  Please consult experts before settling down to a strategy.

1. Self Funding

No brainer… this is nothing but you  spending the money you have.  Either your parents gave you a lot of money or you have made enough in your own career that you can “afford to lose”.  Depending on how much you have,  you may go up to validating the idea enough to quit your job and take the idea fully.  But it is unlikely that you will be able to build a complete business with your own money unless it is a traditional zero-risk business (if there is any like that).

An example would be, if your dad owns a retail grocery shop and you want to start an online wing of the shop.  Probably the existing business (read your dad) may fund this idea.

2. Debt

Debt is nothing but taking loan from someone or some institution to run your business.  It could be an individual, bank or some other.    Typically they give you loan  for the interest they can make.

The Unsecured Loan is something you don’t have to attach any of your property to get the loan.  But such unsecured loans are very few and will come with a lot of terms and conditions.  Your friends and family may give you such unsecured loans but you may want to think twice to take such loans.

The Secured Loan is something that you will show some security to get the loan. It could be any fixed asset (land, flat, gold etc.) that you own.

People who lend you unsecured loan may pay some attention to your idea and its viability.   When it comes to Secured Loan,  lenders pay least attention (if any) to your idea and business viability.  You don’t need to convince them about your idea but they should be convinced about the value of the fixed asset you are attaching.

Debt is a great option if you think your idea is zero-risky.  For example, if you are planning to take a franchise of Pizza Hut in an upcoming  multiplex+mall in a class A city,  how wrong you can go?

Be aware that there are regulations about debt vs equity ratio you should maintain in India.  For example, you cannot fund your company with 100% debt.

3. Friends & Family (& Fools)

This is the option  you go to your friends and family asking for investment.  You will give part of your company (equity) to them in return of their investment.   The biggest debate in this round is how to set a price to the equity?  If someone gives you Rs 10,00,000/- (Ten Lakhs),  should you be giving 1% of the company or 10% ?   All that you have is an idea that is yet to be  validated and the business may not even be in sight.   There are creative ways to arrive at this number like convertible note, but they are too advanced for the scope of this post I want to limit to.

Theoretically, you have to convince Friends and Family about your idea and business for them to invest in your idea. But in reality,  most of them invest because of their trust on you and your team than the idea.

If your idea is an innovative one, you are advised to involve friends & family even if you can put in a lot of your own money and/or debt.   Besides reducing your personal risk, involving those friends and family who may be seen as experts/experienced in the area of the idea will build credibility to you and your idea.   Imagine you going to an investor or a potential customer saying CEO of a product company has invested  25 Lakhs in your business!   That  25 Lakhs endorsement from an industry leader is worth much more than what 25 Lakhs can pay you for your operations.  It is absolutely a smart thing to give a much bigger equity to such people for the same amount of investment they make compared to a novice making the investment.

4. Angels

Angels are individuals who regularly (supposedly) on lookout for investment opportunities in new ideas and businesses.  Typically they are experienced professionals who may have made enough money.   Each individual angel will have his/her reasons why they want to invest. The reasons range from “staying connected to evolutions”, “giving back to echo system that supported them when they began”, “to make more money”  and many more.

A quick search on internet reveals that there are many “angel groups” at national and city level in India and elsewhere.  An angel group is just that, a group of angels.  The advantage of these groups is that they are more organized and you may not have to pitch for each one of them in the group separately.

The downside from my personal experience is that it is  difficult to convince a group about a new idea.  One black-hat in the room can spread the negativity even before the idea starts flying.  Rarely we find ideas that any group of more than 5 will instantly like and get convinced.  Unfortunately,  the one who is completely sold and excited typically do not talk as much as the one who quickly found “a reason” why it may not work.  Almost all the ideas initially will have more than one reason why things may not work.   My personal advise is, avoid group settings early on and until the idea is validated somewhat and a prototype is built.  While this group mentality is true even for F&F and other options, Angel groups is probably the first time you encounter a group setting.

The best thing that can happen to you at Angel round is if you can convince one industry leader in your space about your idea.  If that person writes a check to you,  he/she will also take you to others. We call him/her a “lead investor”.   Others will be less violent at you to dismiss your idea looking at who endorsed it earlier.

5. Seed Funds

Seed fund are mini VCs.   These are institutional investors just like VCs.  Once your prototype is done and small revenue started trickling, you are ready for a seed fund.   There are a plenty of seed funds and quick search will reveal a few.  I do not want to give any links and promote any.

These funds will pool money from their sources and invest in multiple startups.  They have diligent process to evaluate ideas, teams and business before they decide what they want to invest in.  They are also good at giving a value to your company.

Seed funds typically invest much smaller amounts compared to VCs.  Typically  Rs  0.5  to 1.5 Crores is what you may expect from them.   Some of them may have an appetite to go beyond 3 Crores too.

Seed funds expect to exit early (say 18 months) unlike VCs who can wait for many years.

In summary,   Seed funds come in early, invest small amounts and would like to exit sooner compared to VCs.

6. Venture Capitalists (VCs)

Venture Capitalists come with much deeper pockets compared to Seed funds.   They can invest large amounts like Rs 10 Crore or even 100 Crores and more.  They take a larger chunk of your company.   They can wait for longer duration say beyond 5 yrs to break even. Sometimes they come back and invest more second time around if your company does really well.

First time when VC come around and makes the big investment, people call it Series-A.  The subsequent investment is called Series-B.

Typically  the company gets an authoritative value when a VC puts in the money.  Say if a VC puts in Rs 10 Crores and takes 25% equity,  your company is valued at Rs. 40 Crores.   All you founders and other equity holders can now  compute how rich you are given your own equities in the company.  Your equities do mean something from then on!

(I welcome any corrections or additional information in comments to make it more helpful for the beginners)

Posted in Uncategorized | 7 Comments

Why did you name it ‘Lean Startup’ Eric Ries?

If I have to bump into Eric anywhere, that is what I want to ask him.  Trust me, it’s not just an academic nitpicking.   I am witnessing some practical disservice this name is doing to the wonderful and pragmatic approach, that is, Lean Startup.

Startup:

Lets take the second part ‘startup’ first.   Any conversation about lean startup invariably has to start with a disclaimer that ‘startup’ is not referring only to the real world startups working from garages and like.  

Eric’s definition of a startup is :  

‘A startup is a human institution designed to deliver a new product or service under conditions of extreme uncertainty.’

He elaborated how a startup situation can exist inside an enterprise, an NGO or anywhere else.     

There is nothing to disagree with Eric on his views about startups.  However, since the word startup has been stereo typed, cinematized and hence internalized with a narrowed view, why stuck with that?

Lean:

“I know lean, don’t think I need another book on Lean” I heard this multiple times when I suggested ‘Lean Startup’.  

Lean Software Development’ is the concept introduced by Mary Poppendeck in 2003.   

Lean is a borrowed concept from Toyota to identify and eliminate waste within manufacturing process. Mary Poppendieck has cleverly applied that to the software development life cycle.   It’s all about finding what constitutes waste in SDLC and how to eliminate it.

By the time ‘Lean Startup’ came out, this ‘Lean Software Development’ was already well established.     If we dig deep, these two ‘Lean Startup’ and ‘Lean Software Development’ do not contradict each other.    Lean startup helps us not to waste our resources. Agreed.   But the approach and tools given have so little in common.  Or, knowing everything about one doesn’t make us proficient with the other. 

Then why ‘lean’ in the title?

The ideas that ‘lean startup’ brought out (or highlighted) are   ‘leap of faith assumptions’, ‘minimum viable product’, ‘customer development’ etc..    These are the ones that must have made  out to the title. 

-Akki

Oct 20, 2013

 

 

Posted in Agile, Corporates, Startup | Tagged , , , , , , | Leave a comment

My journey as an entrepreneur

YAssume TeamWell, some of you know this already. Few of you suspected but didn’t ask me.  (But most of you don’t care anyway).

My entrepreneur  journey has ended on July 31st 2013.   We relieved everyone from  Makemydabba.com, shutdown the site and vacated the office.   Some of us have already got jobs and are settling down.  I personally am still on look out for my next adventure (ok, lesser than an adventure this time) (This as of Aug 30, 2013).

The next question would be what happened? MakeMyDabba was supposed to be a success right!   I am hoping to document all my learning from my startups in some form some day, but the following is the quick summary for the enquiring minds.

MakeMyDabba :  This is indeed a success as far as the idea is concerned.  We applied all our learning from our earlier startup YAssume and ensured we will not repeat the mistakes.  Went to market within 3 weeks of deciding to start MakeMyDabba.  Validated both supply and demand of the business.   Partly validated the operations model  but there were still lingering questions about operations at the time of our exit.   But the main reason why we had to shutdown was that we ran out of cash to run the operations.   We failed to secure enough funds to reach the traction needed to attract institutional investors.  We founders have already expensed all our own funds and friend & family during our first startup YAssume.

YAssume :  This is a highly technical product targetting sales people.  We nailed down some complex problems,  devised few  original algorithms we can be proud of.   But our focus was not on business for very long.  We did not engage ourselves with our true users of the system.  We did talk to people but they were not the true audience of our system. We mistook the smart-ass opinions to be the feedback needed for the system.  We made steep turns and big development cycles re-orienting the system multiple times within 2 years.  At the end we had a huge complex system but not many real users vouching by it.  We had  to shut it down and move on in Jan/Feb 2013.

What Next ? No secrete, I am truly disappointed.  We were a great team in comparison to many around.  We manage to solve almost all the problems we encountered during our journey.  But we failed in the business.

I have no regrets though.  If the clock is turned back, I will do exactly the same,  that is leave the safe job and try a startup.

I personally lost quite a bit of money on these startups. But learned a lot.  I probably would have got couple of MBA degrees with the same investment and time.  Would my learning from these startups match to those MBAs?  Have to wait and watch.

Thanks

-Akki

(Sep 2, 2013)

Posted in Uncategorized | 5 Comments