All posts by Neil Harwani

Interested in movies, music, history, computer science, software, engineering, management and technology
Uncategorized

Reasonably right decisions – Humanoid framework for AI

Leave a comment

I have a perplexing philosophical question. Currently by using our intelligence and common sense we are able to reach reasonable security. Buy authentic, check promoters/owners, check architecture, feedback, surf, read, sites, notes, work, code, etc, this gets us to a reasonably right decision.

But technology is exploding. Beyond a point of complexity we won’t be able to manage, understand and take reasonably right decisions. In enterprise architecture this new area has to come of keeping complexity manageable so that humans are under control. Otherwise we will not be able to manage the complexity.

Artificial intelligence is now going to catchup. We have Google Now everywhere. New rules, framework, architecture, code, etc integrating with internet are needed. Just like old time network and operating system security.

New framework for and from the values of humans. Hypothesis, laws, framework, rules, architecture and code seems to be the right order. It has to cater to emotions, intellect, biology, philosophy. We have been talking AI for long but now AI is on the cloud and in the internet connecting to everything.

Humanoid framework for AI. This sounds and feels better.

Uncategorized

Software memories

Leave a comment

Many people around us work in tough situations and professions like defense, police, offshore oil rigs, etc. We should have a project where we download their thoughts onto a computer like backup of computer. If they pass away their family will miss them, we cant bring them back but we can definitely give them a computer with the persons’ thoughts, memories, life events, etc and it will give them company just like we have autobiographies.  Something to think about.
Its like an autobiography but with nice virtual reality type feel. One level above a book and pictures.

Innovation, Internet, Social Media

Integrate and evolve with digital, not resist and disrupt

Leave a comment

We hear lot of thoughts by people worrying about integrating with digital world like cloud, internet and so on with respect to privacy. Humanity is all about evolution and not about resist and disrupt.

What we need to do is embrace the digital world and find best practices and standards to move forward. Digital is not going anywhere. Its only going to be more and more around us. Rather than fight and disrupt, lets embrace it and integrate with it. Its just not about data and cyber security, its about what to store where, how much access, when, governance, what to do incase of data breach, backups, best practices, standards and so on.

Lets evolve with digital. 🙂

 

Information processing

One version of truth

Leave a comment

With so many systems, sub systems across the world in web, dark web and internal networks we have various versions of truth in each system. Truth in these systems may or may not be exactly the same version or even be fully/partially right in all/few/one system. We have search across systems but we don’t have a refined framework to find the right version of truth relative to time. With the proliferation of multiple systems, more and more systems going online, more sensors and mobile devices what we need is truth finder search.

It should be able to search across systems, create various states of truth across time and finally give via differential the truth w.r.t. given time.

Culture, Innovation

Race to be leaner, lighter, faster – limits will hit us

Leave a comment

Whether jet engines, aircrafts, IT outsourcing, manufacturing or services there is a race to be leaner, faster, better, higher quality, etc with lesser resources – humans or machines or steel or fibre. In itself its an oxymoron or irony though not always. It actually stops innovation at times.

Let us look how:
Innovation and peak performance happens when a system/entity (human or machine) works at optimum load. Pressure equals force divided by area. If you decrease the area, pressure increases. It causes more stress per unit area. Whether new entity can tolerate it now due to advances is a different stream of research but definitely stress per unit area is increasing. For aircrafts its pressure or stress per unit weight, for humans its stress per person as its assumed people will work with higher efficiency in lean/agile mode.

A fundamental rule which we ignore here is that systems/entities/people will work optimally within evolutionary biology/theoretical physics limits only. We can’t change them. Like if you want to occupy an area you need specific number of people/police/army/robots, this rule doesn’t change. You can augment this by technology but can’t change rules 100%. Extrapolate to innovation/optimum peak performance, its not possible always to be more and more lean, lighter, faster and get better quality and innovation. We need to follow nature’s limits, nothing goes faster than light in this universe.

Disaster relief, IoT, Mobility, Sensors

Software innovation around physical profile of a person

Leave a comment

With so many IoT, mobile, computer, industrial, etc. sensors around us and the number only increasing on daily basis there could be innovative things we could do to solve some of the big problems we face today. There are law and order problems around missing people, criminals, fugitives, finding people and their current condition during disasters, etc. where at times it’s very difficult to get information about people.

Using software to process information from the fused array of sensors available from IoT, mobiles, GPS, etc. we could build profiles of targeted suspects in terms of various attributes like:
1. Physical appearance using mobile networks – CDMA/GSM to scan them
2. Mobile usage related attributes
3. Surfing patterns
4. Preferences around shopping, etc.

This fused profile from various sensors does present challenges around privacy but it also has huge scope in terms of law and order, security, disaster and risk management, etc. which can also do lot of public good.

Imagine during an earthquake, flood, etc. as soon as the networks start coming up we are able to find our friends and relatives along with their present location and condition. Won’t that be great? 🙂 🙂 We could simply have a mobile app which lights up people on our screen in our network who need help.

Continuous interaction, Disconnect fix, Innovation, Onsite Offshore Delivery, Software Delivery

Continuous interaction and collaboration rather than fixed time and schedule meetings

Leave a comment

Working over the years in the Information Technology industry I have learned that face to face interactions lead to success in delivery and innovation which are the important areas in this industry.

But this is the result of imperfect interactions from a distance. For this let us analyze what is a successful interaction:

  1. Successful interactions for good delivery and innovation are continuous in nature and not just fixed hours per week/day/month
  2. If we restrict ourselves to a very fixed schedule say every week or month we come with a very specific preset agenda and miss out on human emotions, behaviors, motivations and so on which also have a big role in success for delivery and innovation
  3. Face to face meetings add on top of the set agenda an aura of informal interactions around friendship, relationship, behaviors, emotions, motivation, what we like / dislike, etc.

To augment this from offshore for countries like USA & Europe and remove the disconnect better way would be to have continuous interaction.

Connect boardrooms, developer areas, testing team rooms, agile rooms, finance rooms, operations room via video conferencing and audio with a view of full team and area on live and continuous basis. Once you do that you remove the fixed agendas and schedules of meetings which cut out the soft side of human behavior, etc. which is only visible in a face to face conversation. This disconnect of fixed schedules and fixed time meetings can be overcome by continuous interactions. We already have continuous integration for delivery teams and AGILE as a model for delivery. Taking it to the next level would be continuous interaction between offshore and onsite.

Information processing, Internet, Universe, Visualization

Visualization

Leave a comment

Visualization in the digital world is an important area often ignored by techies. Visualization in traditional sense of user interface / user experience / mobility for systems and applications has now long been considered an important area.

But another interesting area to look at in visualization would be internet and understanding universe both – yes I am linking both. With petabytes of data/information now on internet and machines with GBs of RAMs and Octa-core CPUs now common place there is an information overload. If you compare the information being generated every year by internet and computers it’s not something that probably a human being could go through in his/her life. Same is the case with the universe, every year we are finding new areas of unexplored universe with new galaxies, stars, planets, etc. An example: The Great Attractor

To understand and process so much information for a normal human brain at current capability does not seem possible (debatable but that’s my view point).

What we need is a little help from visualization and using something like a dimensional / visualization helmet of virtual reality type. Fighter pilots have helmet mounted displays and we need something similar that processes lot of information from internet and universe which could come from multiple dimensions say 5/6+ (more than four dimensions is difficult for us to understand: time, x, y and z axis, etc.)  and parses through petabytes / zetabytes of information and presents it based on normalization for  our senses.

Culture, Excellence, Innovation

Formula for Successful Innovation Driven Companies in IT industry

Leave a comment

There are many formulas to build a successful company:
1. Passion
2. Great investors and board with good mentors
3. Great product
4. Niche skills and services
5. Great legacy
6. Recruit the best, train them, mentor them, give them good career growth, etc.
And so on

But one that I have observed and learnt is another one. We won’t go into the discussion about the earlier ones given above but let’s look at another way to build excellence.

Excellence and innovation is driven by freedom, creativity, knowledge, skill and right intervention by management with positive re-enforcement, promotion of good attitude and removing the bad apples at the right time.

One of the primary enablers for this is less internal friction and a culture to deliver excellence, promotion of passion to perform and learn without unnecessary hindrance.

One of the best ways to do this is to find people in your core teams to go into the non core functions over time willingly and on their own. These people can build a culture that appreciates the problems of the majority core function employees.

An example, if you over time promote in any Information Technology company your core development/product team members via training, education, etc. into Finance, HR, Project Management, Support, PMO, Legal, Sales and other departments you will build a company that appreciates the problems that the core function faces and they will help/enable them which in turn results in innovation and culture of synergy with excellence.

This is not the only formula out their to build a great company but one of the many ways. Only when we perform as one unit do we succeed most of the times, don’t we?

Architecture

Architecture design

Leave a comment

One mistake many make while building architecture for any application / system is to build the architecture using best/latest technology in isolation.

Architecture design has to be mapped to the use case and requirements.

Areas listed below and more should be considered while making design for the architecture:
1. NFR (Non functional requirements)
2. Security
3. Groups/roles/etc.
4. Integration
5. Use case
6. Information processing, etc.

All of the above and more should be considered upfront and consideration should not just be given for the best/latest technologies. Unless for example we consider how many users will use the system, which type of information exists, what are it’s updation rates, etc. who will access the system with what rights, what scalability is needed, how fast will the system change, what are the uptime requirements, etc. we can’t build the right system.

All of the above should be covered right from design and appropriately data should be partitioned, security design should be built, scalability should be thought of and issues like groups/roles/users, identity and access management should be mapped to the use case or the requirement.