Sunday, 19 March 2017

Blockchain (Part 1)

Blockchain. Bitcoin. Distributed Ledger. Are you also one of those who are intrigued by these new terms that are making rounds world over? If yes, don't worry you are not alone and part of approx. 99% of world community who is yet to understand what exactly it means. Even the 1% who say they know it probably do not completely understand it.

Let me try to explain. Blockchain is the underlying protocol of Bitcoin. Distributed ledger is an important component of Blockchain. Bitcoin is a currency that no single country owns. What? Yes, Some unknown person/community/organization named Satoshi Nakamoto had come up with ground breaking idea on Bitcoin in 2008 and then it just took off. People interested in Bitcoin idea formed a group and created the Blockchain network for exchange of Bitcoin crypto currency between network members.

The Bitcoin network participants own the currency which is purely in digital form that has few unique characteristics. Network has capacity to generate new currency to participants based on their contribution to the network activity called as mining. Once it is generated, it becomes part of owner's Bitcoin wallet. The owner can then transfer it to other participants as per his or her wish. It is analogous to the way we have our physical wallet with govt. issued currency in it but, different than online bank account. The participants don't know each other since all have hexa numerical identity on the network.

Blockchain is more generic and broader concept. Bitcoin is one use case of Blockchain. Other features of Blockchain are - pki cryptography and hashing. PKI cryptography makes sure that only authentic owner of Bitcoin (or any other digital asset that is getting exchanged over network) will be able to spend them. Hashing makes sure transactions are connected to each other and are tamper proof. Some transactions are grouped together to form a block and these blocks are connected to each other. Connection of Blocks gives it the name Blockchain.

Distributed ledger helps remove intermediaries. All participants have copy of all transactions happening over network in chronological order. That is why we say nobody owns Bitcoin and all are equal.

Blockchain is democratizing the way people interact and transact with each other rather than any single company providing the platform and earning huge profits out of it. It is going to be threat for many successful businesses such as google, facebook, uber, amazon etc.

Hope you are now better informed on this new emerging concept. We will cover more on it in next part. Keep reading!

Thank you,
MintStat Team

Saturday, 17 September 2016

Artificial Intelligence (AI)

Welcome back. This is an overview blog of AI.

Artificial intelligence is intelligence built in machines or computers with which such machines would take appropriate actions in a complex situation similar to a human being acting after thinking. AI is short form used for Artificial intelligence.

Research in AI started in mid of 20th century and has gained more focus in recent years. Earlier, it was more of theoretical research due to limitations on computing. With advent in computing power, AI has become more practical and hence more interest from different organizations including Google, Facebook etc.

Computer playing chess is one example of AI. Recently, AI is being used to create chat-bots to help customers of any organization. Online customer may not be able to detect who is sitting on other side - whether it is real person or a AI program.

Per our research, AI is still in nascent stage of development. But, it will definitely evolve in next 5-10 years such that AI will become omni-present in all aspects of human life. There is currently fear that AI will take jobs away from humans but, this fear is half true. Yes, AI will take away manual-repetitive jobs from humans but, it will create new jobs where humans will be required to program them and maintain them.

AI involves computing subjects such as machine learning, deep learning, programming and mathematics such as statistics. The goal of any AI project is to process either images or natural language text or audio or combination of these - the way we humans interpret them. Based on training data sets, AI machines take appropriate actions in live situations. Every situations output can be fed back to system so that AI machines become more intelligent. Again similar the way a baby learns.

Hope this short intro helped you to get started with AI. More details in next blogs. Have a nice day!

MintStat Team

Friday, 4 September 2015

What is Not Analytics?

Analytics has become most misunderstood term these days. It has become fashion to mention analytics as feature of any solution. I would like to clarify what is not analytics?

1. Analytics is not the traditional rules based system.
2. Analytics is not the traditional BI reporting.
3. Analytics is not the traditional formulae based templates. 

 Then what is analytics?

1. Analytics is to find hidden patterns.
2. Analytics is to predict events.
3. Analytics is to apply statistical techniques.

Input data to any analytics solution is mostly unstructured. The data is huge. The data arrives at comparatively high speed.

Hope you agree with me.

MintStat Team

Sunday, 26 April 2015

HR Analytics - A force to reckon!!!

Welcome to MintStat blog after a long break. We will now be more regular in this aspect.

Today we want to discuss about HR Analytics. Analytics has been the buzz word in industry. How can HR be left behind from this relatively new technology phenomenon? HR dept's main function is to analyze employee related all data and activities i.e. from recruitment, onboarding, appraisals, compensation till exits.

Most of the organizations still do this analysis manually. If data is small, then it can be very well done by a small team in few days. But, what if data is huge? are there any better ways to do it?

Yes, with more and more research and tools in analytics field, one can write few algorithms with open source tools (such as R) to quickly get insights and help in quick decision making. So if you want to measure effectiveness of recruitment ads in different media channels over the years or you want to measure effectiveness of recruited candidates vs real performance, analytics can help you. The past data would be needed to create good prediction models for recruitment function alone.

Analytics can help you classify employees in different buckets based on performance and accordingly select reward and retention strategies. Based on various text inputs employee has given in company forums, text analytics can identify sentiment of employee.

Employee satisfaction surveys is big area for HR analytics as well. If your organization is interested in HR analytics, please get in touch with us.

MintStat Team

Monday, 17 February 2014

What is Big Data Analytics? Where do we use it?

The current buzz word is Big Data Analytics. Lets understand what it means. It is a three letter word. Big-Data-Analytics.

Big - Where does Big Data Analytics starts? How Big data should be so that it categorizes as Big Data? Big is a relative term. Earlier size of Mega Bytes (MB) use to be Big compared to Kilo Byte (KB). Then Giga Bytes (GB) became bigger than MB. Now, Tera Bytes (TB) is considered as Big data. Soon TB will be considered as smaller and Peta Bytes Or Zeta Bytes will be considered as Big Data.

Data - Here Data means any data. Not only the data that is traditionally stored in databases but also, the data that is present in web, forums, emails or files. Some data may reside within organization boundaries and some data may be in external public sites. Data stored in database is generally known as structured data while data present in forums such as facebook or twitter is known as Unstructured data.

Analytics - Analytics is to analyze the (Big or Small) data and come up with reports and findings that will help management to take right decisions. One can use ready-made tools or can write custom programs as per organization needs.

There are many real world applications of Big Data Analytics. Few examples are:
1. Enhance Retailing - Based on Buyer's profile, past experience, community recommendations and current trend in sales, one can present products to Buyer when he\she is online or in-store thereby increasing conversion rate from just visitor to customer.

2. Financial Services - To reduce frauds, to reduce risks and to cross sell products based on vast transaction data, Financial Services can leverage big data analytics to their biggest competitive advantage.

3. Tourism Services - Travel firms, Airlines and Hotels can use vast amount of real time weather data, world events, ticket sales and reviews to increase effectiveness of their marketing campaigns.

If your organization is looking to implement Big Data Analytics, contact MintStat. Happy Analytics!!!

MintStat Team

Sunday, 3 November 2013

Diwali - A Festival of Lights!!!

Season's greetings from entire MintStat team.

Diwali (or Deepavali) is a Festival of Lights celebrated worldwide by Hindus to spread cheer and joy. It is a five day celebration of human life and relations. The festival is to forget sorrows, worries (~darkness) and enjoy the good things, happiness (~light).

So, if you want to know more about this festival, just drop in a mail to us!!!

MintStat Team


Saturday, 27 July 2013

How to track Quality of your IT project?

Quality is what every project sponsor want from project execution. Important point to note is that quality is relative term. For some zero defect is quality, for others acceptable number of defects is also quality. Zero defects present only in ideal world and our experience tells us that no IT project can be delivered with zero defects. So, there should be some practical but, stringent goal for quality.

Also, achieving quality goal should not add extra cost to overall project budget or schedule. To make sure quality goals are met and no extra cost is incurred, following quality parameters should be tracked:

1. Number of defects  - First set maximum defects goal for your project. You can decide this based on industry standards. Track this parameter at end of each phase of project. e.g. Check number of defects detected at end of Analyze, then after Design, then after Build. The number of defects should reduce in each later phase.

Defect prevention meets help reduce defects in each consecutive phases. So make sure you identify root cause and action item to avoid defects properly. 80/20 principle helps in defect prevention meet.

2. Functionality coverage - Once requirements are frozen, make sure all functionalities are covered during each consecutive phase of project. Any missed functionality identified during testing will add extra cost to project as team will have to re-do design and coding.

Requirement traceability matrix helps cover 100% functionality coverage during each phase and can avoid above extra cost.

3. Effort spent on reviews, rework and testing - Make sure sufficient efforts (hours) are planned at start of project for reviews, rework and testing. Keeping it very low or very high can impact quality or cost of your project. Track this effort again at each phase of the project and identify variation / action items, if any.

Use charts and graphs to easily track variation and set control limits for this variation.

We feel if above parameters are tracked properly and proactively, one need not worry about quality. Please contact MintStat if you want to make sure your project achieves all Quality goals.

Thank you,
MintStat Team