Wednesday 25 March 2020

Need for effective Data Management Practices

Data Management Practices
If you ask what the most precious thing for an enterprise is, everyone will provide the same answer, Data. Why data is so important now because we can generate so much of it through various means and we are using it to make critical decisions that will make or break an enterprise or its efforts to become successful.

Since data is this much critical, having an accurate version in the format we want is the most important action that would decide the quality of decisions an enterprise makes.

In the remainder of the article, let's see a few best practices for Data Management.

In the remainder of the article, let's see a few best practices for Data Management, starting with what data management?

The Concept:

Data Management spawns right from the planning of what needs to be collected as data to retiring of the collected and used data. Every organization should be very clear on what they need as data so that it will help them in what decision making and other usages. This clarity is very important so that they can devise an implementation strategy to collect data, develop those tools for data collection and make sure it is cleaned, transformed and ready to use state.

There should also be a data retiring plan which is part of the data management. The retiring plan is when an organization has used that data and that data is not relevant anymore, then what should they do with that data? With the concept of what is data management out of the way, let's see a few effective data management practices in the remainder of this article.

With the concept of what is data management out of the way, let's see a few effective data management practices in the remainder of this article. 

The Best Practices:

As explained in the concept section, data management starts with planning.

Planning means, what to collect, from where to collect, what processing needs to be done after collection, how to use the transformed data and when to retire that data?

All these questions can be answered only if an organization has clear goals on where they want to go? the purpose, that purpose will drive the efforts on their decision making which in turn will drive their data management practices.

By doing this will limit the data collected which means an organization will collect only data that is needed to them. This will also greatly improve data quality, which is the next best practice we are going to see.

Need to have quality data to make accurate decisions, right?

The first step is to collect only what is required. The second step is to make your team members aware that data quality is a very important process that they need to adhere to. 

They should be well trained on those steps that are used to clean data in your organization.

Most of the time data come into an organization in an automated way, hence, it is very important for the team to go through them, find imperfections and correct the same before sending that data upstream to analyze and generate reports for example. 

The team should also be aware of figuring out stale data, data that has lost its value and of no use to the organization and remove them periodically.

Few other important steps are to keep checking for data duplication, missing data, wrong data, wrong data format, misleading data types, etc., 

Along with these, data security, which is storing data in a secure way and make that data readily accessible by the teams that need these data should also be a key priority in your data management process.

Solution Consultant

Wednesday 18 March 2020

COVID-19 epidemic impacts mental health, Here is how ML can help

Covid-19 affect mental health , Machine Learning can help
How Machine Learning can help in Mental health crisis

Covid-19 has claimed lives, closed borders and house arrested millions of people around the globe. From toilet paper panic buying to livelihood uncertainties the effects of the Coronavirus epidemic are only growing day by day. At this point, there is no escape from the effects of the virus. Whether we are infected directly from it or not, our way of life has changed dramatically over the last few weeks.
While many of us are getting adjusted to the new norms of life, one least looked area of the impact is mental health. Business owners are worried about the uncertainty and collapsing market. Employees are worried about not getting paid and not being able to support their families.  Some cant get work from home option and so they have a tough time finding alternatives for babysitting.  

We can’t deny the fact that this outbreak has created stress in a lot of people. Fear and anxiety have taken over. While young adults are passing memes and discussing discomforts of being at home, adults either afraid about their own health or that of their loved ones.  Being isolated and at home has increased sleeping and eating disorders.

When tragedy like this strike, we meet our friends and family, go to prayers or play sports. We hug, touch and comfort people. With the Covid-19 outbreak, we can’t gather in churches, meet our friends, play sports, have drinks, have gatherings or have physical contact to comfort ourselves / others.  On the other hand, we are constantly bombarded with news about the challenges and hardships of the outbreak.  Mental health is going to one of the priorities for all of us as we sail through this hardship.

Machine Learning in diagnosing mental health

AI has taken over several industries including different verticals of the health industry. Machine learning a subset of AI is showing good signs in the subjective diagnosis of mental health. ML even supersedes human capabilities in accurately identifying the disorders.

There are no blood tests for mental health, and often humans can miss out on the cues of the patients such as the words they use, which can be symptoms of mental health. This is where ML can play a major role as they are good at identifying subtle cues, changes in day to day speech, etc. This is exactly a team of researchers from the University of Colorado Boulder are doing.  Using Machine learning in psychiatry, they created a speech-based mobile app that can categorize a patient’s mental health status as well as, or better than, a human can.

Researchers from the World Well-Being Project (WWBP) analyzed social media with an AI algorithm to pick out linguistic cues that might predict depression. Although these are at the early stages of addressing the tremendous need for the mental health care of this situation, it provides a good starting point. It also enables us to exploit our resources and create such applications to address the needs of the crisis.

Srivatsan Aravamudan - Sri

Senior Solution Consultant

Thursday 12 March 2020

How robots, drones and AI can help combat outbreak

AI and Robots are making a huge difference in the current Corona Virus crisis

Let’s face it corona is not the first epidemic the modern times have seen, and it’s not going to be the last.  Fueled by the need to support the elderly population Singapore has already deployed technology in aiding healthcare professionals in hospitals. These ‘smart hospitals’ incorporate machine learning algorithms in managing patient visits to robots helping patients in various ways.

For example, they can help with direction, delivery of medicine and are tested to perform complex tasks such as changing linens or lifting heavy equipment. The robots are also linked to the fire alarm system to ensure the safety of the patients and can navigate through obstacles on their path. In Hubei province, the epicenter of the epidemic, a Chinese tech company Cheetah Mobile Inc is offering automated services, in the hope of reducing the burden on front-line medical workers.

Reducing the cross-infection in outbreaks.

Remote consultation is already in use which can reduce the spread. However, in growing panic over outbreak patients rush to hospitals making it difficult for medical workers. The sheer number of patients coming in through the ER would make it difficult to provide medical care to everyone in time. It can often result in medical workers working longer periods and exhaustion. During the outbreak, the actions that need to be taken for medical care, such as taking temperate and asking a set of questions for diagnosis can be a repetitive and draining experience for the medical workers.  More importantly, working with a large group of infected patients also increases the chance of cross-infection among medical workers.  If the medical workers turn to patients, it will only increase the burden on other medical workers to fill that space.

With the current advancement in robotics, we can overcome this situation by placing robots in the ER rooms. The robots can ask the questions for diagnosis, take temperature,  prioritize the cases based on machine learning algorithms, and can isolate high-risk patients immediately. Robots can then be used to dispense medicines to these patients.  Robots can be used to sterilize the room with human intervention. More so robots can be sterilized easily before meeting the next patient.  Thus, minimizing the interaction of medical workers and improving the efficiency of medical care during an outbreak. Robots don’t get tired, and they can work perform the same task consistently.

Beyond hospitals

During outbreak cities get locked down making it difficult to send supplies and medicines to the affected areas. Self-driving vehicles are being used in Wuhan to deliver medicines. Chinese e-commerce company has been using autonomous vehicles to move packages to hospitals.
Drones aka the flying robots are also being used effectively in the affected areas. Preprogrammed drones are used to sprinkle disinfectants, take thermal imaging for diagnosis in addition to patrolling the public spaces.

AI in alerting and monitoring

If you are wondering whether AI is used to monitor the outbreak situation, you are right. Metabiota and BlueDot were used to track the initial outbreak of the novel coronavirus. BlueDot's solutions track, contextualize, and anticipate infectious disease risks. BlueDot alerted the Corona virus threat even before WHO and CDC issued public warnings for the disease. It did that by monitoring online content and processing huge amounts of data using NLP.

While all these high technologies can aid us during the outbreak,  we also need to be vigilant and responsible during the outbreak. Please follow  CDC’s instructions for keeping yourself and your family members healthy, such as washing your hands and avoiding travels.

Srivatsan Aravamudan - Sri

Senior Solution Consultant

Friday 6 March 2020

Machine Learning revolutionizing these 3 fields

Machine Learning revolutionizing several fields 
Artificial Intelligence is an inevitable part of our future and one of the most utilized practices of AI is machine learning. Every business in every industry wants to use some aspect of machine learning to get ahead in the competition. The growth of machine learning is quite exponential. It’s also quite possible that you are already using some form of machine learning in your everyday life even without realizing its presence.

Machine learning is simply a computer program performing some form of cognition like that of the human brain. It is this aspect that gives the edge to machine learning applications from ordinary computer programs. Using ML (machine learning) computers don’t need explicit commands or instruction to perform a task that humans(us) want them to.  They achieve this by working on the sample data and making predictions or decisions to carry out an assignment. Every application of ML is revolutionizing industries in the ways we cannot imagine. In this article, we will investigate the top 3 applications of ML.

Personal Assistant

I say Hey google at least 5 times a day, and I know many of us are in close contact with Siri, Alexa, and Google.  Personal assistant’s success lies in its ability to learn your pattern of queries, collecting and retrieving answers more suitable to your needs.   You can also ask your assistant to set alarms, read your schedule, send directions or even order something online.

With natural language processing (NLP) which is based on machine learning, assistants can be programmed to process and analyze human language input.  The first step in this process is to let humans communicate with the system in their own language. The second aspect involves in personal assistant understanding the commands and performing the required tasks.

According to Grand View Research, the voice and speech recognition market hit $9.12 billion in 2017. And It’s expected to grow at a compound annual growth rate of 17.2% from 2018 to 2025. The bottom line here is that the market for personal assistant driven by machine learning is growing.  With more companies joining in to make their services accessible with voice recognition ML-powered personal assistants are going to take over most of our everyday personal digital chores.


By 2020, 30% of all B2B companies will employ AI to augment at least one of their primary sales processes according to Gartner. AI/ machine learning has the potential to eliminate the time-consuming, manual tasks of sales teams and help them spend more time with customers. With ML, we could automate account-based marketing support with predictive analytics. This would support account-centered research, forecasting, reporting, and would be able to recommend things like which customers to upsell first.

Machine learning technologies are good at pattern recognition and enables sales teams to find the highest potential new prospects by matching data profiles with their most valuable customers. Most AI-enabled CRM applications do provide the series of attributes, characteristics and specific values that pinpoint the highest potential prospects.  This helps the sales team to save thousands of hours a year in selecting and prioritizing new prospects.

According to Salesforces’ latest State of Sales research study majority of guided selling adoption will accelerate based on its ability to rank potential opportunities by value.  Machine learning-based guided selling will be based on prescriptive analytics that provides recommendations to salespeople of which products, services, and bundles to offer at which price.  


According to Deloitte. Machine learning improves product quality up to 35% in discrete manufacturing industries and is only expected to grow up. McKinsey predicts that 50% of companies that embrace AI/ML over the next five to seven years have the potential to double their cash flow with manufacturing leading all industries due to its heavy reliance on data. The bottom line is ML is providing insights on improving shop floor productivity thus maximizing the product quality and production yields.

According to a recent survey by Deloitte, streamlining the inbound supplier quality is the priority in the manufacturing industry. Machine Learning /AI technologies can consume a lot of data including audio and video making them the best quality analysts in the production floor. Systems equipped with Machine Learning capabilities are already preventing the breakdowns by quickly identifying the anomalies in the production equipment or processes. Even better AI-enabled systems are predicting the patterns with respect to failures using sensors enabling the industry to focus on preventive measures to address that.

Machine learning AI-powered systems are particularly helpful in automating complex tasks consistently to improve throughput, energy consumption, and profit per hour.


While these are some of the industries ML has revolutionized there are several more that needs to be talked about. In business, Machine learning has a wide range of uses. In fact, most of us interact with AI/ML in some form or another daily. From the mundane to the complex, AI/ML is already disrupting virtually every business process in every industry. As ML technologies proliferate, they are becoming an imperative for businesses that want to maintain a competitive edge. Talk to us to understand how we can support in finding that competitive edge for your business.

Srivatsan Aravamudan - Sri

Senior Solution Consultant

Tuesday 3 March 2020

Building an App for Business (3/5) - Application architecture

Application Architecture redefines the success of the business app

Welcome to another article in the series of building an app for business. If you haven’t checked the previous articles on this series here, they are : 
Building an App for Business (1/5) – Starting Right with Ideation

Let’s come to the topic of this article, what is application architecture and why it’s crucial for the business to pay attention to it. 
An application Architecture is the process of defining the framework of an organization’s application solution and match it against business requirements. 

Architecture ensures the application landscape is scalable, reliable and manageable. It also defines the what major business functions to support and how to manage the data.

A financial organization offering trading or monetary transaction through its application decided to revamp the application for adding new features based on the feedback from the customers. However, they have launched the changes without thinking about the architecture of the application. Users got stock trying to access their stock account details or during the process, the application reported several bugs and in the end, the organization lost its customers to the competitors who provided a more seamless experience.

According to Gartner application architecture determines how well the application performs during the volatile situation. Poor application architecture not only reduces business agility but also dampens performances and reduces vulnerabilities.

Business Value

Adopting agile methodology can really help in delivering application which can scale, quickly change according to user requirements and reduce time to market. However, having a clear business value aligned to the organization’s vision and customer focus is what makes the application successful. Thus, the application architecture needs to be mapped with key drivers and should be able to communicate the changes and its impact.

Flexible emergent architecture

Adopting to rigid complex architecture upfront as in the BDFU approach can really do harm to business apps. A business app needs to scale and change according to changing customer expectations and increasing competition. An emergent architecture combined with an agile approach provides just enough structure but enough room for improvement and changes.


Application architecture is going to be one of the key decisions for the organization deciding to take a plunge in application development. Development decisions must be collaborative development architects, application leaders and developers. Always consider how the application development infrastructure impacts the outcome, set clear objectives, knowledge-sharing platform and opportunities. If you like to strategies your application architecture write to us today.

Srivatsan Aravamudan - Sri

Senior Solution Consultant

Related Posts Plugin for WordPress, Blogger...

Share Buttons