Tech Fiction

I have been working at Dataiku for 2 years and a half now. What an amazing ride it has been so far! The company has grown from 50 people in one office in Paris to 200 across the world, from a startup to an actual company. I feel that things are getting more and more interesting.

- How many times have I watched that movie? Not enough.

I was personally lucky to get the opportunity to move from the Parisian "mothership" to the blooming London office. It has been very enriching to work on the other side of the Channel. Again, I would like to thank the founders, my colleagues and my clients for the trust they have put into me thus far.

- Fond memory of a team dinner near our office in Aldgate.

As this blog post is the third instalment of a series, this year I would like to take a different angle.

Most of you already know that the Data Science / Artificial Intelligence (AI) market is maturing and that we received more money in the bank to accelerate our growth. But these are not the deeper reasons why I joined Dataiku instead of another AI company.

1. Human potential

Four years ago, I was a young consultant. I did not know how to code. I knew next to nothing about "Data Science", "Machine Learning" or "AI" (that was not even a term back then). Apart from the business strategy slides I was making, I knew very little of actual data science. During four years, partly at school and mostly on the job, I learned to become a data scientist.

Having said that, I am not saying that everyone must become a data scientist. While this is a job I really enjoy, it is not the only one and it can only exist within a much larger ecosystem. We should all remember that data science is a speciality within AI, which is itself a speciality within the "IT crowd". As a matter of fact, data scientist are now being broken down into smaller sub-categories.

What I really like about our software, which makes it unique, is that it is not only for data scientists. It is designed for all data professionals to use it, whether you are a chief executive, a data engineer, a business analyst or a data scientist.

By being inclusive of all roles, our tool creates the conditions for everyone to reach one's potential.

2. The importance of practice

Working on many real deployments of AI projects has taught me one thing: while the theory is the foundation, there are many lessons you can only learn through practice. From a Python notebook prototype to a real-time predictive API running in production on a Kubernetes cluster and delivering business value, there is a (reasonably) long path.

What we are trying to do at Dataiku is to develop the different paths to deploy AI projects. I like to compare it to a network of hiking paths, let's say our beloved European GR. We make the physical paths, removing boulders that may be in the way. We put signals so people don't get lost. We maintain them so people can have a smooth and enjoyable hike.

But we are not trying to pretend there is a magic button to go from "Point 0" to "Point AI".

It takes practice, because there is not a single path, and because it is about more than technology. As an organization, you need to bring everyone on board, not just your IT department. To do so, you need to a "hiking toolkit" which comprehends the many paths that exist and which can be adapted to your own needs.

That's what we aim at with our product: helping our clients walk the AI paths.

3. Kindness and passion

I would not be here today if not for the kindness of the people I have met. We are shaped by our environment: friends, family, coworkers, fellow practitioners.

I am very lucky to experience this kindness in the way we work at Dataiku, not just for the challenge but to make each other grow. I am still amazed at how much time experimented people such as our CTO spend to explain their work to newcomers. That inspires me to give back during user trainings, conferences and my daily life.

Please don't take my word for it and check out my colleagues' opinions.

I will not lie and say that everything is perfect. There are many areas to improve, things to change, processes to create. But our collective passion makes it possible to go through the sometimes rough patches of scaling a startup. To me, the key is to preserve that passion with respect to the problems we face.

If that made you curious about us, we are looking for 200 people to join the fun wagon this year! If you have any questions, please send me an email at alex [dot] combessie [at] gmail [dot] com!

I wish you all a very happy 2019,

Last October 2016, I was starting a new job at an uprising French startup called Dataiku: fifty people developing one product (Data Science Studio, a.k.a. DSS) with the will to transform the way companies use data.

- Couscous for lunch with a few colleagues last October!

I was both excited and nervous. As an otaku “data scientist”, I loved the Japanese play-of-word, the software and the vision behind it. There was so much to learn about so many aspects of data, business, computer science, … I was impressed by the diversity of skills of the team and more importantly, their willingness to share knowledge.

Fast forward today, at the edge of 2018, I feel that all of this is still true. So I would like to thank all my colleagues, from interns to co-founders. They have become companions, partners, and friends. Day by day, I feel lucky to work with such a great people, to keep growing and learning.

- Who does not like bad puns with a GIF?

Thankfully, some things did change!  We grew more than x2 in people, clients, and activity in general. In various regions of the world, we are now talking to big companies who had never heard of us a year ago (ping, Gartner Quadrant fans). Looking back without falling into the temptation of vanity, it has been quite an amazing year.

I feel that we are reaching a tipping point, approaching the size where structure matters more and more. In the Paris office, we have even have elected a “comité social économique“! Having worked mostly with larger companies before, it is interesting to participate in this particular arc of the adventure. I am not an unconditional fan of processes but I understand they are useful when done well. I hope we retain our fun and flexibility while we scale up!

Indeed, Dataiku is different today compared to when I joined. But I still deeply believe in our part to play in the world of data. By this foggy end of December 2017, as bellies grow and time slows down, I would like to reflect and re-state the reasons why I joined Dataiku.

1. Data Science is not one job. There is more to data than scientists. Coming from a non-technical background, I am convinced that everyone can use data with the rights tools. Scientific expertise is needed for Machine Learning and Artificial Intelligence, but this is only a given component of what people do with data. 

As a matter of fact, Machine Learning often comes last, after data has been captured, analyzed, and after the business problem has been defined. This requires varied skills, including those of a "typical data scientist” (though I think there is no such thing) but not only. In 2017, the centralized model of data scientists as an elite with the sole privilege to analyze data has shown its limits. Data is for everyone, at different levels of technical expertise. 

This has been one of the initial assumptions behind the development of our software, and I think it resonates with the state of the market today.  Some of our larger clients have hundreds of DSS users, and we see more and more projects where data scientists, analysts, and business people work together on our tool.

2. Data security matters. This is becoming a number 1 priority for all organizations today.  It bothers me that 123 million of personal records can be lost because someone forgot to add a password to their cloud storage bucket. It bothers me even more that this negligence can happen two months after the Equifax leak when 143 million credit records were stolen by hackers. 

These breaches need to stop. Personal privacy is a fundamental right which needs protection. In Europe, the General Data Protection Regulation (GDPR) will be enforced very soon next year.

But this goes beyond data leaks. Everyone working with data should be accountable for the consequences of their work, automated or not. Unfortunately, there is no such thing as the Hippocratic Oath of data scientists. Our work having an increasing influence on people’s lives, there needs to be a form of regulation.

Having a centralized platform for data science is a key step to control how data is used in an organization. Our software is just a tool and obviously, it cannot replace human and organizational safeguards. But it offers a way to enforce these safeguards more easily. In the context of complex open source stacks (think security on Hadoop), this is especially useful.

3. Business expectations are increasing. Data labs have been around in large companies for a few years now. They usually started as experiments and were not directly driven by short-term return on investment (ROI). I see this changing now that data labs have established the proof of their value.

Consequently, two things are happening. On the one hand, data lab teams are getting more budget and visibility within their organization. On the other hand, their management (often CXO level) is increasing the pressure for ROI. It is becoming more and more important to define a business case for each proof of concept (POC).

However, there is still a gap between the initial POC and its industrialization. This can require significant time (6-18 months) and introduce new risks: what if the team in charge of industrialization uses different tools than the data lab? How to monitor the performance of Machine Learning models over time? How to manage versioning and rollbacks?

After having industrialized several client projects with DSS, I am convinced this is where we bring the most value. While it is easy to make a successful POC without DSS, problems start to appear as you approach industrialization. Awaiting for this, no business value can be extracted… Having a tool to secure and accelerate industrialization is crucial to making the business promises of data science come true, and fast.

I am very happy and grateful to continue to be part of the Dataiku saga. I am convinced there will be many new quests to accomplish along the way. As a matter of fact, the next act for me will be London, UK! I am excited to join the British office starting March 2018. 

I am excited to share some great news with you on this blog. I have just joined Dataiku, a fast-growing startup developing a novel kind of Data Science platform. This is an important new step in my life, after spending a few years in the consulting business and training as a Data Scientist. I would like to give you some explanation as to why I joined them and why I believe in their vision and product.

(Big credits to French magazine Le Point for this cool picture)

My own vision for the future has not changed: I am still convinced that Big Data & Analytics technologies, especially Machine Learning, are a positive force of transformation for our economy and society. In my years at Capgemini Consulting, I have seen amazing applications of such innovations, both in France and worldwide. We have gone a long way since the beginning of the Big Data phenomenon in 2011-2012 (see chart below).

We are now at a tipping point where Big Data & Analytics are no longer buzzwords. The value has been clearly proven ; companies are investing both on people and technology. Now is time for real large-scale applications. The journey is 1% finished! I believe Dataiku is at the forefront of this revolution, leading the charge to make Big Data & Analytics promises come true. 

I have been using their product Dataiku DSS for a year before joining them, and I loved it from the very start. I still remember my sense of wonder when I first used it as a young Data Scientist: finally, I could spend less time on basic data preparation, stop worrying about Big Data infrastructure & dependencies and focus on what is important: developing great data products. All this without ever losing the “power of code” and building upon open source solutions: Python, R, Spark, etc. As I saw other Data Scientists in my team discover and love this tool, I realised how powerful it was. Simply put: after becoming their biggest fan, I finally caught the ball and joined them:

All in all, I think Dataiku DSS is an amazing solution to the three key challenges of Big Data adoption and productivity:

Empowering everyone no matter the expertise: Data is not only for scientists; everyone should be able to use it, from executives to analysts. Collaboration is the key feature to transform organizations with data.

Navigating a complex technological landscape: companies need to combine the best of both worlds: their legacy systems and the sprawling ecosystem of new data technologies (a.k.a. Technoslavia). Organizations need an integrated solution to manage all these systems in a transparent way.

Building data-driven products: Proof-of-Concepts are a first step, but real benefits come from their industrialization in production environments. This is where an end-to-end approach is needed.

Two weeks ago, I participated in my first hackathon, organized by ENSAE ParisTech and Microsoft Research. The event gathered 50 students for a 24-hour data science competition, based on data provided by the charity organization Croix Rouge (Red Cross). There were two categories of challenges:

Predicting the daily need at each Croix Rouge center up to 2 weeks in advance

Determining where to install new centers in France (the one my team and I chose)

It was an amazing experience, and I am very honored that my team received the first prize in its category. A picture is worth a thousand words, so here goes the winning team, aka the Data Buyers Club:

I am really proud of the work we accomplished in less than 24 hours, with little sleep and a lot of coffee/soda/pizza/etc. Generally speaking, I was very impressed by everyone’s work and the level of involvement in the competition. There were a lot of interesting models and insights, and I hope it will help Croix Rouge and the people who need it.

Here are some “impressions” from the event and our prize:

We spent 80% of our time on data loading and preparation. Machine Learning was only 20% of the job.

Big Data is hard and it requires a lot of time to handle it. Sometimes it is interesting to turn it into “Small Data”. We chose to drastically reduce the geographical scope of data, so we transformed a 10-gigabyte database into a table that could fit in Excel.

Sometimes, the simplest models are the most efficient. Most of our work was based on a linear regression. We tried Ridge and Random Forests, but it took too long.

Presentation is key. In the world of Data Science, great code is nothing without visualizations and storytelling.

To conclude, I would like to thank the team of advisers at Microsoft Research for their help during the hackathon, especially Xavier Dupré, who teaches programming at ENSAE ParisTech. His blog is an amazing resource to learn Python, algorithms, and Machine Learning.

I realize that I have not given much news since July. The last few months have been quite busy / productive, so I had to take a short break from blogging. I will try to be more regular from now on.

As explained in my previous blog post, I have decided to go “back to school” to become a Data Scientist. Since September 2015, I have been studying Data Science at ENSAE ParisTech while working part-time at Capgemini Consulting. The goal is to obtain a Master’s degree in Data Science by mid-2016.

I would like to conclude, not with a note of self-congratulation, but with a big Thank You.

"If I have seen further, it is by standing on the shoulders of giants."

- Isaac Newton

Thanks to all the researchers, innovators, and entrepreneurs who have developed the theories, tools, and techniques that enable today’s Data Science. It is really a shared community effort, from academia to industry.

As I announced a few weeks ago on my social networks, I have been accepted to the ENSAE ParisTech program in Data Science. Starting in September, I will follow this program as a part-time student while continuing to work at Capgemini Consulting. I will progressively transition from my current management consulting role to a Data Scientist one. I am glad to share this amazing news with you, and I am grateful for all the help I have received to make this project come true. Special thanks to the Big Data & Analytics team at Capgemini Consulting for supporting my ambition!

I have not communicated about this project on my blog, so you may wonder about the motivations behind this project. In short, the main reason is that I love mathematics and computer science. I have been a science nerd fiddling with computers since I was 10, and I am proud of it. Even if I have chosen a business career to start with, I am deeply attached to my engineering roots. Out of intellectual curiosity, I have drifted into finance & economics after 3 years of pure engineering, but I have always had the project of going back to it at some point.

The first trigger came during my internship in finance when I realized how much mathematics and data were central to today's business life. It was in 2011, and the word "Data Scientist" did not have the publicity it now has. Then, during my final year of studies in Toronto, I entered the local startup ecosystem, and the idea continued to grow. The second trigger was a presentation by Peter Suma, co-founder of Applied Brain Research, at a startup course I was attending. It opened my eyes to a new world of possibilities. When I went back to France to find my first job, I was determined to start working and building my skill set in the area of Data Science. I happened to find a great consulting firm with a clear vision of Big Data & Analytics, willing to invest on it, and with a growing team.

I have been experimenting with Bitcoin for over a year now, and here is a short summary of my thoughts about it:

Pros

Bitcoin can drive transaction costs down dramatically, both disrupting the traditional finance industry and creating new business models based on micropayment. Knowing the high fees imposed by current payment intermediaries to merchants and thus to consumers, Bitcoin is a big deal.

Bitcoin does not require much infrastructure investment (near-zero marginal cost) which makes it an interesting option for unbanked populations in developing countries.

Beyond payment, the underlying Blockchain technology can be used to enforce complete contracts, extending its range of applications to law, politics, real estate, etc.

Cons

Even if transactions are publicly logged, Bitcoin is anonymous by design, which creates big regulatory risks and barriers to adoption both on the merchant and user side. The commonly-heard argument "the dollar bill is anonymous too" is wrong. Why reproduce a historical flaw into a feature? The case of Silk Road has shown that Bitcoin can act as an accelerator for illegal transactions and money laundering.

Due to its decentralized nature, there is no clear way to tame the volatility of the Bitcoin exchange rate against fiat currencies. The only solution would be a massive price fixing operation, which is unlikely at this point.

A personal prediction for 2020

Bitcoin will fail as as a mainstream currency, except for some specialized use cases such as international payment. Along with tighter regulations, some companies will build layers above Bitcoin to mitigate its anonymous and decentralized nature. Despite the relative failure of Bitcoin by itself, the tech industry in general will manage to bring transaction costs down. Finally, the underlying Blockchain technology will serve as an example/framework for online contract enforcement, which mainstream applications appearing by 2020.

# Transportation

Uber raised a total of $3 billion in three rounds in 2014, making it one of the most valued tech companies, at $40 billion. Previously, only Facebook had raised that much money ($1.5 billion in 2011). Uber does not communicate much on its internal metrics, but some leaks provide a rationale for this huge valuation.

Lessons to be learned :

Industries that have not embraced digital transformation yet are at a very high risk of disruption

Mobile is becoming the preferred channel for selling goods & services

Uber is a perfect example of the combination of great software, fast operations scaling, and aggressive marketing strategy

Questions :

In light of the recent regulatory backlash, how fast will Uber continue to grow?

How will Uber address the controversies about its company culture?

Will Uber manage to expand beyond personal transportation to become a global logistics operator?

# Mobile Messaging

Facebook acquired WhatsApp for $19 billion to take advantage of its active user base of 600 million people. While WhatsApp current business model ($1/user/year) only generates $20 million in revenue, Facebook will probably explore other monetization opportunities in the long term. Generally speaking, the mobile messaging market has exploded in 2014, with Rakuten’s $900 million acquisition of Viber, Tango’s $280 million funding round, etc.

Lessons to be learned :

Messaging is the killer app of mobile and social media

It is more than just messaging: these apps are becoming platforms for content, gaming and payment

Questions :

What will be the monetisation strategies to justify these huge valuations?

When will the market consolidate?

# Mobile Payment

With the iPhone 6, Apple introduced a mobile payment feature using Near Field Communication to pay in-store and in-app. As for now it is only available in the USA for major credit/debit cards and at select retail stores, but it will expand in 2015. In response, Samsung has started working with Paypal to integrate a similar feature in its future smartphones and smartwatches.

Lessons to be learned :

Apple Pay is the catalyst that will open the latent mobile payment market

Financial services will face a new level of technological disruption in the coming years

Questions :

Will existing payment networks survive?

Which parts of the banking industry are most at risk?

# Wearable devices

All major equipment manufacturers have released or revealed their first line of smartwatches. In almost all cases, they are built as an extension to users’ smartphone, with additional health sensors. Similarly to the smartphone world, the industry is divided into 2 main platforms : Google’s Android Wear and Apple’s adapted version of iOS, with Samsung trying to push its own proprietary platform.

Questions :

What will be the killer app of wearables?

When will the technology mature and the market take off?

Will wearables bring as big a change in our lives as mobile devices did?

# Smart Home

Google acquired smart thermostat maker Nest Labs for $3.2 billion and home monitoring startup Dropcam for $555 million. All major players are getting ready for the explosion of the mainstream market, with Samsung’s acquisition of home automation platform SmartThings, GE’s partnership with hardware invention startup Quirky, etc.

Lesson to be learned :

Following this year of industry consolidation, consumer adoption will accelerate in 2015

Questions :

What will be Apple’s response to Google and Samsung’s aggressive strategies?

Who will be the winners of the upcoming smart home platform war?

# Machine Learning

Google acquired Deepmind Technologies for $400 million, to leverage top talent in the specialized field of Deep Learning. This acquisition brings Google additional artificial intelligence capabilities, with powerful applications such as image recognition, speech recognition and natural language processing.

Lessons to be learned :

Machine Learning talent is scarce and highly valuable

The frontier between academia and the tech industry is blurring

Question :

How will other companies be able to bridge the talent gap?

# Virtual Reality 

Facebook acquired Oculus VR for $2 billion as a long-term strategic bet. While the Oculus Rift virtual reality headset is still at the development stage, we can expect a consumer version in 2015. Meanwhile, Samsung and Google have released prototypes of virtual reality headsets using users’ smartphone as display.

Lesson to be learned :

After decades of development, virtual reality technology is getting ready for mainstream adoption

Question :

Will virtual reality expand beyond gaming applications?

# Drone

Both Amazon and Google announced that they are working on package delivery drones: Amazon Prime Air and Google X Project Wing. This technology could revolutionise last-mile logistics and serve people in remote areas and places with bad road infrastructure. Regulators have not approved it yet, but it is a key proof of concept of this technology. Meanwhile, the consumer market for drones emerged, with new products from Parrot, DJI, etc.

Lessons to be learned :

The commercial drone technology is ready but the market is still untapped

We are waiting for improvements in air traffic control systems and regulatory decisions

Questions :

How will regulators address the terrorism risk?

Will it be a revolution for developed or developing countries?

# Autonomous Car

Google announced that their self-driving car prototype is functional, after more than 1 million kilometers of test on the road. For the moment it has been tested on well-chartered territory in a few american states, including on public roads and in traffic. Despite limitations such as the inability to recognize temporary road signs, this is an important step towards future commercialization. Tests with safety drivers will continue on public roads in more states in 2015.

Lesson to be learned:

Do not underestimate the long term potential for automation

Questions :

What will be the regulatory response?

How long will it take to reach mainstream adoption?

What economic and social consequences on the long term?

# Robotics

Qualcomm demonstrated Zeroth, a prototype of neuromorphic chip that mimics the humain brain. This technology could be the future of  computing and help robots better understand and react to their environment. Compared to existing technologies, it would open up real time applications, consume less power and work without connection to the Internet.

Lessons to be learned :

The future of computing will be inspired by biology

It is only a matter of 5-10 years for robots to become mainstream

Question :

A few words on my personal relationship with Foursquare

Foursquare has been on my smartphone home screen for 3 years now - a long time considering today's shortening technological life cycles.  I have been one of the early adopters of the service, back when it was a simple checkin app. I have over 1000 checkins, 100 tips, and I used to have a ton of badges (back when the feature existed). I even created a few places to contribute to the community.

As an avid life-logger, I love how Foursquare empowers me to remember all the places I visited, and how it creates incentives to discover new ones. I also use the Search feature a lot to prepare my nights and trips with friends. The only feature I have never used much is the social part, because too few of my friends are on Foursquare. To me, the value created by Foursquare is clear :

What about Foursquare’s recent decision to unbundle its app?

Following an initial announcement in May and made official in August, Foursquare split its app into two separate ones:

Swarm for Social features (checkins, making plans with friends, …)

Foursquare for Discovery features (search, review, …)

This decision has caused a lot of controversy lately (see reviews on Google Play or the App Store). In my personal case, despite an initial negative reaction, I adjusted my user flow to use the service for the same purposes as before. But the point is: why did Foursquare decide to make such a radical change of strategy? I see two potential reasons, which are interesting lessons for entrepreneurs:

1. One App = 1 Use Case

There are some underlying trends going on:

As the mobile app space matures, the quality and quantity of apps increases, so that developers are competing for attention

The technology allowing apps to communicate with each other is improving rapidly (deep-linking, etc.)

Due to the nature of the smartphone form factor, there is a limit to the number of features people actually use in an app

As a consequence, app developers need to focus on one use case and a few key features to make the value proposition clear to users. For existing apps, it means unbundling apps by use case (and using deep-linking to make apps work together). There are plenty of examples of this phenomenon: Facebook with Messenger and Groups, Google Drive with Docs, Slides and Sheets, Dropbox with Mailbox and Carousel. To borrow the expression of my favorite VC-blogger Fred Wilson, we are entering the "app constellation" age.

2. Crossing the chasm

Beyond the benefits of unbundling described above, splitting was also a necessary step for Foursquare to « cross the chasm » between the early adopter and the mainstream market. From my perception, Foursquare was built on a early adopter community consisting mainly of life-loggers who used it for check-ins. That approach worked out really well to build Foursquare’s database, since life-loggers acted as real life web crawlers. However, it did not scale very well in terms of user growth and monetization, since life-logging remains a niche market. Moreover, it became a barrier to mainstream adoption, since most users are uncomfortable with the idea of sharing their personal location data. The presence of Social features prevented the Discovery uses from growing.

That was Foursquare’s dilemma: to cross the chasm, Foursquare had to break its life-logging image, risking to alienate its early adopters, to refocus its value proposition around Discovery.

I have been thinking a lot lately about the differences in the process of innovation in established companies and in the world of startups. In particular, why do incumbents fail to innovate and resist disruption? I have several well-known examples in mind: the music industry, the hotel industry, and more recently, the taxi industry.

There are several factors I can think of:

Lateness to realize the seriousness of the threat (systematic mistake)

Fear of cannibalizing established products (this is true of the music industry)

Inability to cooperate with the other incumbents on a common standard or platform to resist disruption (this is especially true of "atomized" industries like hospitality and transportation)

Here I would like to provide an alternative explanation of this phenomenon based on the popular Hype Cycle framework developed by Gartner Research. I think it is a simple but powerful tool to illustrate the pattern of progressive disruption:

Here is the story:

Incumbents are aware of the seriousness of the threat on the long term, and they do experiment with technological innovation. However, these experiments often fail to live up to the high Return on Investment (ROI) expectations. As a consequence, the incumbents decide to focus on the existing core business. Meanwhile, the technology matures and the ecosystem of startups grows. Most fail, but a few succeed and turn into profitable businesses, either through resegmentation or disintermediation. On the consumer side, a critical mass of early adopters emerges and starts reaching out to an early majority.

By the time the incumbents realize the immediate threat, it is too late. We reach a no-return tipping point with two possible paths: "Grow fast or die slow". Disruption has happened.

What is the bottom line here?

Innovation does not bring direct ROI but it is a long term investment

Risk and failure are part of the innovation process

The timing and the scale of projects matter: don't start too small too early

Although disruption starts slow, tipping happens very fast

Hello world,

This is the first post of my new blog/adventure in the land of "Tech Fiction". I want this blog to be a repository of my personal thoughts on technology, innovation, economics, and life in general. I will try to post new items once a month. Length and format will be free.

A few words to introduce myself:

I am a young French engineer passionate about technology and entrepreneurship

I have lived in France and in Canada, studying engineering, economics and innovation

I have worked as a financial analyst for big companies, and held business development roles in startups at various stages

I am currently working in France as a strategy consultant at Capgemini, and learning how to code on the side

Why I am starting this blog:

I have been tweeting about the tech industry for about 2 years @alteralec

I was starting to feel a bit limited by 140 characters 

I am seeing some trends and patterns in what I read and hear around me

I believe in the value of sharing

I hope you will find this blog interesting!

Cheers,