Data Science Round-up: March 20th, 2019

Reading Time: 1 minute

Here’s our round-up of the most essential reads on Data Science from across the Internet:

Data Science is now bigger than Big Data: Data Science is now commanding the hype that Big data used to have, as organisations are looking for new ways to leverage a large amount of data they have to produce insights.

Why Data Science Teams Need Generalists: Data Science generalists have improved autonomy and are empowered to make better business decisions because they have better knowledge of end-to-end business capabilities.

Europe is better for Data Science compared to the US: Europe’s heavy investments in education and infrastructure and education has paid dividends in better skills.

How Data Science is powering Nvidia’s business: Data Science has become the new driver across the computing landscape, encompassing analytics, AI/machine learning, and inferencing

Secrets of a successful Data Scientist in 2018: A report on the most important skills that successful data scientists have and other key information.

9 Applications of Data Science in E-commerce: Interesting applications of Data Science in e-commerce that are driving results.

 

Artificial Intelligence Round-up: March 20, 2019

Reading Time: 1 minute

Here’s our round-up of the most essential reads on Artificial Intelligence from across the Internet:

How AI is changing science: Automation algorithms are assuming the role of a scientist and are crunching numbers at a cosmic scale, but they still lack the spirit of scientific enquiry. We still have some way to go before that can discover physics or mathematics that the brightest humans alive are not able to do on their own.

A Robotic Leg Learns to Walk Without Prior Knowledge: Researchers have developed a bio-inspired algorithm that can learn a new walking task by itself after only 5 minutes of unstructured play, and then adapt to other tasks without any additional programming.

AI generated poetry: A retrained version of an older algorithm has finally learned how to make rhyme and meter work to generate poetry based on old classics. The results are middling, but its a promising step forward.

DeepMind and Google; Controlling AI: Google acquired DeepMind for $600 million, and is now looking to monetise the company. But the company’s founder is resisting attempts with a bid to focus on AI research and producing path-breaking technologies.

5 most promising roles for Artificial Intelligence in India: Roles such as AI architect, AI product manager and more are defining the growth of AI, and offering professionals a fulfilling and rewarding career.

Everything you need to run Deep Learning experiments: Learning by doing is the best way to understand Deep Learning better, and you need some serious computing firepower to run experiments. Here is a list of all the components you’ll need to build your rig.

 

5 most promising roles for Artificial Intelligence in India

Reading Time: 5 minutes

There is a massive demand for skilled AI professionals right now in India, and there is a lack of talent that can fill these roles. 76% of companies feel this lack of talent is hindering their ability to implement AI. For working professionals looking to meet this supply-demand gap, it might be confusing to choose the right aspect of AI to specialise in, owing to the broad nature of this domain. So we’ve put together this list of the most impactful, high-demand roles that are available right now, as seen below:

This is by no means an exhaustive list, but a collection of roles that are most likely to be highly impactful on the growth of the AI industry, as well as offering high growth opportunities for professionals engaging in it.

DATA SCIENTISTData Scientist in AI and ML

Data Science as a domain is almost as broad as AI and ML when it comes to the scope of implementation across different industries. Data Science is a combination of data inference, algorithm building and implementing technology in order to derive insights from a data set. Since it uses a fair amount of mathematics and statistics, there’s considerable overlap with Artificial Intelligence. Data Science is already a domain that has a lot of demand for professionals, so becoming a Data Scientist with a focus on AI would help professionals carve out a lucrative niche in both domains and build a rewarding career.

Data Scientist can be a very broad term that describes a host of functions across a host of different domains, but in this specific instance, a data scientist is responsible for the massive amounts of data that are required to build robust AI models. Their chief responsibility is analysis, that ranges from collecting data to manipulating to providing actionable insights. In most cases, data is unstructured and unsanitized, and the data scientist needs to be able to structure the data to make it processable by the AI algorithm. Any real-world application of AI needs high-quality and relevant data for it to deliver the expected results, and it’s the role of the data scientist to manipulate the data into a machine-readable format.

AI SOFTWARE ENGINEERAI software engineer

Software development is a crucial part of any company’s technology journey. Just as software engineers have been tapped to build various digital products in SaaS, Cloud, Social Platforms, e-commerce sites and so much more, someone needs to build AI products as well. Software engineering in the AI sector has just barely passed its nascent stage as new applications of AI are sprouting up all the time, so there’s a lot to learn, and consequently, a lot of growth potential as well. 

The role of the software engineer is vital to the functioning and management of AI models and applications. Obvious as it sounds, the role of an AI software engineer is to build the AI models and the applications, where the AI architect will piece them together to form a working solution. Software engineers need to be proficient in the programming language of choice, along with a set of routines and functions, called libraries, that are used frequently.

AI ARCHITECT

AI Architect

Computer or systems architecture deals with the specifics of building systems, by defining the functions, organisation of components and the implementation of the system as a whole. As computer applications started becoming complex, systems architecture became necessary to address the complex requirements of modern applications, by integrating the software and hardware components effectively. In Artificial Intelligence, where the complexities are exponentially higher, architecture becomes crucial to build an effective AI model.

This is one of the most crucial roles to sustain an AI model, and the lack of AI Architects actually hampers the ability of companies to execute their AI models. An AI Architect is responsible for translating the companies business objectives to efficient and profitable AI implementation, so it’s a really high-impact role in a high-impact industry. They chart out the structure of the AI implementation and design how the different components of AI work together to deliver a solution.

AI MODEL TRAINERAI Model Trainer

Model Training is one of the core aspects of effective application of AI. Building the model is a challenging task, but model training helps to fine-tune the results to ensure a higher degree of accuracy. Until now, model trainers have worked with relatively simple algorithms, but the sheer amount of data that AI needs throws up a unique set of challenges. Model training is not just about feeding data to the algorithm, but also to ensure that the input data is relevant, free of duplicates and does not contain any other rogue data that could derail the model. 

This role would overlap with the Software Engineer or Data Scientist, but as AI systems mature and become more sophisticated, AI trainers will grow into a specialised role where their sole task is to train AI models with preferred data that improves their overall accuracy and effectiveness. They usually do this by manually feeding the training algorithm with relevant data that improves accuracy and performance.

AI PRODUCT MANAGER

AI Product Manager

Product management deals with building, testing, managing and improving a digital or physical product. Product management includes planning, forecasting, production or marketing of a product at different stages of the product lifecycle. This role would have different responsibilities in different organisations, but the overall gist of the role is what we’ve just described. Product managers are an important part of any company, because products are usually the lifeblood of a company’s revenue stream.

It’s crucial to look at AI as a product rather than just as a supplementary service. The function of an AI product manager, as any other product manager is to be the point of contact for all teams, acting as a single point of contact for all stakeholders of the product. They are responsible for charting out the roadmap for the product and are involved in aligning the product development with the business goals of the company. They also take part in activities such as taking the product to market, collecting user feedback and converting them to feature requests, and much more.

If any of these roles interests you and you are looking for a way to launch your career in Artificial Intelligence and Machine learning, you can consider Great Learning’s Post Graduate Program in AI and ML that teachers learners industry-relevant skills through hands-on projects.

I am from a non-technical background but I found it quite easy to follow the videos: Madhan Deverajan, PGP-CC Alumnus

Reading Time: 3 minutes

Learning once and applying it for the rest of your life is a long-lost trend. With technologies evolving continuously, ‘Learning for life’ is the new mantra to stay relevant. In the case of Madhan, our PGP-CC alumnus, a professional with 15 years of experience, upskilling was the only way out to excel in his existing job. Read on to find out how Great Learning’s postgraduate program in cloud computing helped him to, in his own words, ‘manage projects effectively’.

Why did you take up the PGP-CC program by Great Learning?

I am an IT project cum service transition manager. I basically manage end to end IT service transitions from various clients and customers to our company. I work for UST Global, which is a service support company. If there is a requirement from a customer for certain IT services like cloud, be it level 1 or level 2 support services, I manage end to end transition right from setting up the process to the transformation from the customer to our company. I ensure that a cloud support team is set up and we are able to deliver support services to the various IT customers. I have over 15 years of experience in handling IT projects. The primary reason for taking up the cloud computing program is that just certifications are very narrowly focussed. I went through the brochure of Great Learning’s PGP-CC program and realized that it touched upon various topics right from development on the cloud to basics of setting up cloud infrastructure. I needed an overall understanding of the cloud so that I would be able to manage projects effectively. Now, I feel much more confident in understanding the problems. I took up this program to perform better in my current role. I had 3-4 choices, but I found this program as more comprehensive than other online programs. So, I opted for this.

How is learning from the program helping in your current role?

All the projects that I did during the program were hands-on projects so that gave me an understanding of the various components of a cloud system and now I know what configurations should be made and what are the checks that need to be performed for running a particular cloud service. Earlier, I used to completely rely on the technical staff or the architects, but now I am able to suggest better options instead of doing something the traditional way.

Are you Azure or AWS certified?

One of my long term plans is to get Azure or AWS certified.

What did you like the most about the program?

1.Curriculum relevance.
2.Assignments, labs and projects.
3.Mr Nirmalya. He was truly amazing during the mentoring sessions that I had attended. I am from a non-technical background but I found it quite easy to connect the dots and follow the videos and sessions with him. His inimitable and fantastic delivery of technical sessions in an understandable way was extremely suitable for an end user like me.

Was the program office cooperative?

Mrs Ekta from the program office had been great. Without her persuasion and mentoring, I would have given up on the course as I was finding it difficult to catch up with the labs and projects midway through the course. She had been open to feedback and was prompt in getting back with answers to our queries and also had been very patient with us. I was given the support and time to catch up with the rest of the group. I am taking some time out to practice the syllabus all over again and possibly plan for further studies in about 6 months with Great Learning.

What advice would you like to give to the aspirants?

I wouldn’t say that the program is very tough, but it is demanding. My advice would be to stick to the deadlines and schedules and work through your practical lab classes. Do not miss even one single session and make sure you watch all the videos. I am able to see the value of the lab sessions in real-life cases now. Even if you are not a very technical person, videos will help you in understanding everything step by step and connect with your peers. Make sure you are not alone that is what even Great Learning’s program office recommends. Connect with fellow learners and get the best out of the program.

Weekly Data Science Round-Up – March 13, 2019

Reading Time: 1 minute

Here’s our weekly round-up of the most essential reads on Data Science from across the Internet:

8 mistakes to avoid in Data Science interviews: Candidates must avoid common mistakes such as just having theoretical knowledge, forsaking domain knowledge, not paying attention to communication skills and more.

Data Science in Retail: Data Science provides the foundation for a company to integrate its key business functions to deliver the best outcomes for themselves and their customers.

97,000 Data Science jobs vacant in India: Data Science jobs have increased at over 45% compared to last year with more companies adopting data-driven decision making.

5 Data Science jobs at Target India: 5 Data Science Job Openings for Data Scientist, Item Data Analyst, Data Engineer At Target India that you can apply for right away.

Why Data Science is a great career option for freshers: Data science will be a rewarding career option because of the higher potential for growth in the future and better pay, along with the huge amount of demand for qualified data scientists right now.  

2019 Report – Career Opportunities in India for Analytics and Data Science: This report explores the types of career opportunities that are available for data scientists across experience levels, geographies, skill sets and more throughout India.

Weekly Artificial Intelligence Round-Up – March 13, 2019

Reading Time: 1 minute

Here’s our weekly round-up of the most essential reads on Artificial Intelligence from across the Internet:

1. Top 5 Successful Startups led by women: A list of startups in India that are led by women, featuring pioneering startups such as Rivigo, Zestmoney, Bash.ai and more.

2. Google uses Machine Learning to predict floods in India: Google’s Flood forecasting project helps to predict floods in Patna, using machine learning to predict the likelihood of flooding.

3. Candidates take recruitment test powered by AI: 534 candidates have received job offers from companies, after taking a recruitment test that was unsupervised by humans, called an Auto-Proctoring test.

4. AI powered ‘Bolo’ improves reading skills by 64% in UP: Google launches Bolo, which helps children improve their reading skills by encouraging them to read aloud, and verify it through a speech-recognition feature.

5. 5 must-haves on your Artificial Intelligence Resume: To land an AI job, you’ll need to know these programming languages, concepts of machine learning, and some engineering science.

6. 3 Machine Learning models used by Spotify to recommend music: Spotify uses Convolutional Neural Networks, Natural Language Processing and collaborative filtering to power their flagship recommendation feature.

6 Common Interview questions for Cloud Architects

Reading Time: 3 minutes

A cloud computing architect is responsible for the structure of an organisation’s cloud infrastructure. The architect builds the foundation on which the front-end and back-end platforms communicate with each other working in tandem with cloud-based delivery systems and the cloud network, all of which comprise cloud architecture. Since the role is critical to the functioning and success of cloud services, the role comes with a high level of responsibility, as well as reward.

Professionals usually transition to a cloud architect position after a few years of working as a cloud engineer or a software engineer, or by doing a comprehensive program in cloud computing. Here is a list of common questions they might face in their cloud architect interview:  

1. How do you connect on-premise applications to cloud services?

You can connect on-premise applications to cloud services through Hybrid integration. Traditionally this has been done with VPNs or FTP for loading data, but they’re not very efficient for integration.

You can choose a good cloud provider like AWS/Azure/GCloud so that you have more control over it, compared to a public cloud set up. You will then have to set up an encrypted channel so that your on-premise apps can communicate with your private cloud platform.  

2. What should be the strategy for application migration to the cloud?

There is no single right answer for this question as there are multiple answers depending on the circumstances. Here are the different strategies:

Re-hosting: This can be done by redeploying applications to a cloud-based hardware setup while configuring the application host accordingly. This is quick and easy while sacrificing scalability.

Re-platforming: You can choose to run the application directly on the cloud vendor’s platform. This can be done without changing the app’s core architecture. While it has the advantage of backward compatibility, the underdeveloped PaaS infrastructure means that common capabilities are sometimes excluded.

Repurchasing: This might come at a high cost and cause lock-in with the new vendor, but one option is to junk the old application and purchase a compatible SaaS platform.

Refactoring: This can be a resource-intensive exercise because it involves overhauling how an application is architected. This is usually treated as a last resort when new features need to be added, or if the services need to be scaled or to improve the overall performance.

3. What is Serverless (AWS Lambda)?

AWS Lambda lets users run code without having to manage servers, so it’s called a serverless service. It executes code only when needed and scales automatically, from a few requests per day to thousands per second. This allows users to pay only for the compute time consumed – Users can run code for different application types without any administration.

4. What is the problem in Lambda (Serverless) implementation?

Vendor lock-in:  When the application is hosted on a serverless platform, porting it to another platform is cumbersome due to compatibility issues in terms of supported programming languages. It also means that the user will have to cede control of their hosted application to the vendor.

Long-term challenges: Since Lamba is a Function-as-a-Service, it calls multiple functions if it takes too long to execute a task, which is resource intensive and ineffective.    

5. What is the use of API Gateway?

The API gateway is the entry point for an API for a group of microservices. It handles protocol translations which makes it suitable for microservices that make use of multiple APIs. They allow devs to design the application structure depending on the use case. The API gateway can be used for:

  • – Authentication
  • – Security enforcement
  • – Cache management
  • – Load balancing

 

6. Where to use NoSQL databases as compared to traditional relational databases?

You should use NoSQL database if:

  • Need to handle a large volume of data that is structured/unstructured
  • If your business requirements make you follow development practices such as agile sprints, quick iterations, and frequent code pushes
  • Prefer object-oriented programming that is easy to use and flexible
  • Want to leverage efficient, scale-out architecture instead of expensive, monolithic architecture

————————————————

These are some common questions that are asked in interviews, but this is not an exhaustive list. Your interviewer might ask you questions that are specific to the company or the role and you need to be prepared for that too. To give yourself an advantage over others, you can choose to pursue a cloud program that offers you comprehensive and industry relevant skills.

I received excellent support from the program manager: Sujay Kumar, BACP Alumnus

Reading Time: 3 minutes

I pursued the BACP (Business Analytics Certificate Program) to get a headstart in my organization’s analytics team.

I am from a manufacturing background. When I got a strategic project management role in my organization Nalco, I knew that I needed to upskill to analytics quickly. They were setting up an analytics unit in Pune and hiring across a wide experience range, so I needed to learn analytics to become a part of the team. A program like BACP has immense brand value and I knew that it would help me a lot to substantiate my analytics knowledge to my company.

My current role demands that I spend more time in the office so an online format was perfect for me.

My current role is at an upper middle management level which comes with its set of challenges and travel time. I wasn’t sure if I would be able to allocate enough time to studies in a full-time program, so I opted for an online format. As per inputs and reviews from other BACP participants, I was convinced that BACP was going to be quite a different experience with its micro-classes (study sessions in small groups of 5). I also realized that at the end of the day, whichever program format you opt for- blended, online, full-time, you will be driven by your interest and intent to learn. Self-study is an extremely critical component in any program.

Is BACP the right program for me?

I must have asked this question to myself at least a 100 times before enrolling as I was hard-pressed on time. Honestly, I did my market research, but most of the courses or programs required me to have prior statistical knowledge or programming experience, etc. Due to my background in the manufacturing industry, I had some statistical know-how in my arsenal, but I was averse to coding and programming. Feedback and reviews from alumni actually convinced me the most to go ahead with BACP. Post completion, I have started working on analytics projects but our projects are very basic unlike the sophisticated models built in other companies. It’s a manufacturing company predominantly driven by sales. I mostly use linear and logistic regression in my current project.

My biggest challenge during the program was managing my study time.

The success of any course does not depend on the course curriculum, the faculty, or any other components. It mostly depends on your dedication and hard work. After a 12-hour day at the office, spending an additional 3 hours even on the weekends seemed like an uphill task. Stretching oneself has its limitations but I knew that in order to do justice to the time and money I had invested in the program, I will have to get over these time constraints and establish a studying rhythm that would last for the entire duration of the program. I received excellent support from the program manager who kept my spirit high even during the low days when I just wanted to give up. Dipayan Sarkar, my mentor was very helpful and encouraging. And even today, I am in touch with him.


Lifelong Learning as a mantra is embedded deep in my system.

After my graduation, I was sure of one thing that I would always keep myself updated and upskill from time to time. Over the past 10 years, I have pursued several programs, at least one program every two years. While others may disagree, but the days of learning once and applying the learning forever are behind us. Learning new things gives me the confidence to catapult both my personal and professional growth.

My advice to aspirants wanting to pursue this course is to be serious about it.

I somehow feel that doing a course just for the sake of a career transition will not be successful all the time. The candidate should be really serious about learning something new. From a learning perspective, it should be more about the process rather than the result. One more thing to be clear about is the projects you will be working on once the program is completed. Analytics is not just about building fancy models from day one. It will take a long time (even 4 to 5 years) to become a proficient data scientist for anyone entering the field without experience.

Is an AWS Certification enough to launch a cloud computing career?

Reading Time: 2 minutes

Cloud Computing is a fast-growing domain in our digital economy because it has upended how organizations store and retrieve their data and digital services. AWS was an early entrant into this field, as a cloud computing platform that offered a scalable tool that addressed the infrastructure requirements through on-demand cloud computing platforms to individuals, companies and governments, on a subscription basis.

To improve operational and delivery efficiencies, companies turn to AWS to host their applications. AWS has firmly established itself in the existing market as a top-rated cloud computing service. That’s why being certified in AWS would be a valuable asset if you are trying to build a well-rounded cloud computing resume.

The AWS certification course is designed to ensure that you gain an in-depth understanding of the Amazon Web Services’ objectives and principles. The most common objective of taking this course is to learn to design and scale AWS cloud implementations. Having to be recertified every 2 years also ensures that the skills of the participants are always relevant and up to date.

Top 3 reasons to choose an AWS Certification

1. Being certified as an AWS professional demonstrates to potential recruiters the
depth of information and your commitment to learning cloud computing.
2. AWS certified professionals are consistently listed among the top paying
info-tech certifications worldwide.
3. Over 68% of enterprise respondents reported using AWS to run their cloud applications. There are more opportunities compared to other platforms.

So should you choose to pursue an AWS certification? The simple and direct answer is yes, but is an AWS Certification enough to boost your career? That’s where cloud experts would disagree.

Choosing just AWS Certification might be a good first step because the cloud computing arena is constantly evolving and changing. Microsoft Azure and Google Cloud are also strong contenders for being leading cloud providers and are offering young professionals a competitive edge.

To truly launch your career in cloud computing, you’ll need a comprehensive program that covers all the aspects of the cloud, that gives a well-rounded understanding of the fundamental concepts. It needs to equip you with the right skills and tools that help you build innovative cloud solutions. An extensive program needs to offer certain benefits to students, such as:

1. Learning to apply concepts and work with tools and platforms across industry-relevant use cases.
2. Learning from experienced cloud computing industry practitioners.
3. Personalised mentorship from cloud architects working in leading companies.

In comparison to just AWS Certification, a Post-Graduate program in Cloud computing offers you holistic cloud skills that will enable you to succeed in your career. Experts at Great Learning have put together a comprehensive Post Graduate Cloud Computing Program that consists of cloud environments such as the ubiquitous AWS, along with Google Cloud and Microsoft Azure. Spread over 6 months, PGP – Cloud Computing is an extensive course has helped people in putting their careers on the fast track to success.