Why are AI use cases not going live? MLOps bring an answer

Over the last years, many organizations have been investing substantially in data and analytics. The objective is to become more data-driven, become a tech style organization. Companies willing to go further than just symbolically profiling the organization, invest in AI to go from descriptive analytics to predictive and prescriptive analytics. This requires a solid data and AI governance program, IT infrastructure that makes all data readily available in a so-called data lake, and piloting of the organization through a carefully selected key performance indicator portfolio. It has been widely documented however that the most important hurdle is the change to a culture that embraces agility and experimentation. In fact, it is the humans that need reskilling. As a consequence, training programs have been launched and large organizations can now boast about their hundreds of use cases created by interdisciplinary teams which are shared on an internal repository for further development and innovation. The hard question comes next: what is the return on these huge investments? Why are so little AI use cases in production and where is the generation of tangible value? There seems to be a gap that needs to be filled and MLOps are bringing part of the answer.

Before going into MLOps, let us take one step back. It has always been a brain teaser for the software development community to find the best methodology for project management. It started with the waterfall approach, introduced in the 70’s by Winston Royce. This linear approach defines several steps in the software development lifecycle: requirements, analysis, design, coding, testing, and delivery. Each stage must be finished before starting the next and the clients only see the results at the end of the project. This methodology creates a “tunnel of development” between gathering the client requirements and the delivery of the project. For many years, this linear approach has been the cause of tremendous loss in resources. An error in the design stage or the clients changing their mind required rebooting the development process. Furthermore, engineering teams were clustered in different stages (developers for coding, QA teams for testing and Sysadmin for delivering) which created frictions a fertile ground for communication errors. This is one of the reasons which led to a new methodology which started around 2001: the agile approach.

Agile principles have infused the software engineering culture for more than 20 years. It has endowed companies with the ability to adapt to new information rather than following an immutable plan. In a fast-changing business environment, it is more a question of survival than a simple change of methodology. Now, companies put customer involvement and iteration at the heart of the software development process. They bring together engineers with complementary skills within teams coordinated by product managers to regularly release pieces of software, gather feedback and adapt the roadmap accordingly. This was a true revolution, but it was not perfect: there was still a gap between software development and what happens after the software is released, also known as operations. In 2008, Patrick Debois and Andrew Clay fill this gap with the DevOps (contraction of development and operations) methodology. By bringing all teams (software developers, QA and Sysadmin) together in the development and the operations processes, waiting times are reduced and everyone can work more closely, in order to faster develop better solutions.

Back to today, what can bring DevOps today in the era of artificial intelligence? The needs are the same: companies are looking for methodology to develop and scale AI algorithms to generate value and reap the benefits of their investments. Data leaders recently began to investigate the benefits of the Devops methodology. However, machine learning and AI algorithms have a peculiarity that drastically differentiate from traditional software: the data.

Data is everywhere and has become a tremendous source of value for companies. The recent advances in fundamental research and the democratization of machine learning through open-source solutions has made artificial intelligence accessible for all. Data scientists are one of the most sought-after profiles in the current job market as they promise to be the key factor in unlocking the value of data. But in the same way that software developers needed Devops methodology to maximize their productivity and scale software development in controlled and secured environments, data scientists need a framework to develop and scale AI-powered solutions. Since those solutions are different from traditional software, they need to be managed accordingly. Therefore it is essential to use Devops practices, but data leaders also need to acknowledge the singularity of using data within software that makes decisions autonomously. This is where Machine Learning Operationalization (MLOps) comes to rescue.

MLOps is a set of practices, bringing Devops, machine learning and data engineering together to deploy and maintain ML systems in production. This is the missing piece which allows organizations to release the value contained in data using artificial intelligence. With formalization and standardization of processes, MLOps fosters experimentation but also guarantee rapid delivery, to scale machine learning solutions beyond their use case status. Once the solutions are in production and consume new data, monitoring predictive performance is key. Universal outperforming ML solutions for specific solutions simply don’t exist, hence organizations need monitoring predictive performance in real time. MLOps helps monitoring this performance and acts in case deterioration due to concept drift occurs. The automation of the collection of lifecycle information of algorithms, that is tracking what has been recalibrated by whom and why, allows improving the learning process and reporting to auditors if required. Hence, accountability and compliance issues can be addressed.

While most data training programs focus on the elements of machine learning, statistics and coding, and work on use cases in a sandbox environment, MLOps principles are not yet covered extensively. Furthermore, business leaders invest in AI without fully understanding how to create an efficient development and operations environment for their data teams. Filling the gap between data and operations is not straightforward. The complexity of ML algorithms, often considered as a black box ran by data scientists who are supposedly the only one in the company to understand what they are doing, separates others from the development process and creates another gap between AI and business.

MLOps does not only concern engineers, every stakeholder of data-based solutions should be involved. The revolution of artificial intelligence is undoubtedly happening now, and all those who intend to be part of it will have a role in creating and running MLOps processes in their organization. Future data leaders should acquire basic MLOps skills in their training programs to remove the harmful and unnecessary boundary between business leaders and engineering teams around data-related topics.

Authors:

Regis Amichia, Data Science Lead, Foxintelligence

Jeroen Rombouts, Professor, Essec Business School

Published at Sat, 10 Apr 2021 10:05:09 +0000

Report: Indian Government’s Initiatives to Boost Big Data Education

The Indian government has been proactively taking initiatives to boost big data education in India. Taking cognizance of the imperativeness of big data, the Ministry of Science and Technology and Ministry of Education have taken relevant steps that can foster big data and artificial intelligence in higher education. These steps have been formulated to sustain and accelerate the digital transformation process across India. Meanwhile, several initiatives have been started, to support the education of big data and data analytics.

The Ministry of Science and Technology under the Big Data Initiatives Division has started a Big Data Initiatives Programme, under which financial aid will be provided to support the research and development of big data across India. Financial aid will also be provided for support to establish a Center for Excellence in Big data analytics, predictive technologies, cybersecurity, etc. Additionally, the programme supports national level conferences/workshops/seminars/brainstorming sessions, etc financially and will support the In-house programmes for Faculty and UG/PG/Doctoral students. The programme focuses to enhance the 5V i.e. Volume, Velocity, Variety, Value and Veracity, of big data education amongst students.

In 2016, the Comptroller and Auditor General of India have initiated Centre for Data Management and Analytics (CDMA) for synthesising and integrating relevant data for the auditing process.

The government agency National Institute of Transforming India (NITI) Aayog has also strategized to integrate technology with education. It has partnered with various international organizations to bolster big data education in India. For example, Microsoft is helping the Andhra Pradesh government in predicting dropouts of students in school education. This initiative, so far based on machine learning and analytics has identified about 19,500 portable dropouts from the government schools in Vishakhapatnam district in the year 2018-2019. Additionally, NITI Aayog has also partnered with the National Association of Software and Services Companies (NASSCOM) and has created an AI-based module to bolster artificial intelligence education in India. The module was launched by NITI Aayog under Atal Innovation Mission (AIM) and contains activities, videos and experiments to bolster academic knowledge of big data analytics and artificial intelligence amongst the students.

Indian Government’s Investments in Big Data Education

India contributes a high number of graduates, especially from a technology-related education background to the global job market. Henceforth, the country holds an important place in the global education sector. Remarkably, India has one of the largest networks of higher education institutions in the world which leverages investments and mergers from the government. The country has the largest population of about 500 million people in the education seekers age group between 5 and 24 years. The education sector in India was estimated at US$102.7 billion in 2019.

Technology-related studies contribute to the biggest number in the Indian education industry. Big data is one of the growing sectors increasing its education ground in the country. Big data has become an integral part of every radar. Therefore, the education sector can no longer afford to remain insulated from the technology connection. A lot of educational institutions in India are using new technological tools to improve pedagogy. Today, the country’s students have access to data and resources available anywhere in the world, thanks to big data and its developments. In the big data education industry, analytics tools and platforms plays an important role. Analytics covers a wide family of problems mainly arising in the context of a database, data warehousing and data mining research. The Ministry Human Resource Department ensured that several national tech universities in the country have set up AI centres for educational, and research and development.

These universities include the Indian Institute of Technology in Kharagpur and Madras and the Indian Institute of Information Technology Design and Manufacturing in Kancheepuram. In the interim budget (2019-2020) alone, the Indian government allocated INR93,848 crore (approximately US$13.15 billion) to the education sector, and a part of it also goes to tech-related studies. According to a document released by India’s Policy Commission (the National Institute for Transforming India- NITI Aayog) titled the National Strategy for AI, AI can potentially solve for quality and access issues observed in the education sector. Private investment in educational technology, broadly defined as the use of computers or other technology to enhance teaching grew 32% annually from 2011 to 2015, rising to US$4.5 billion globally.

India is currently among the major big data analytics markets in the world. Analytics Insight estimates the Indian Big Data Analytics industry to reach US$18.9 billion by 2025 from US$5.5 Billion in 2020, growing a CAGR of 27.8%. It will account for 32% of the global analytics market the same year which will increase tremendous scope in education and employment. India is witnessing high growth in the big data analytics industry, due to its large resource pool of technically skilled English-speaking population. The education sector in India remains to be a strategic priority for the government. The government has allowed 100% Foreign Direct Investment (FDI) in the education sector through the automatic route since 2002. According to the data shared by Department for Promotion of Industry and Internal Trade (DPIIT), the total inflow of FDI in India’s education sector stood at US$3.24 billion between April 2000 and March 2020.

In 2020, the Indian government has multiplied its investment in its innovation program known as ‘Digital India’ to INR3,063 crore (US$477 million) to support research in AI, IoT, big data, cybersecurity machine learning and robotics. Under the project, the country aims to make internet access available to over two lakh villages by 2019, promoting e-governance, e-banking, e-education and e-health. In Budget 2019, Finance Minister Nirmala Sitharaman announced that the government will ensure industry-relevant skill training for 10 million youth in India by building skills in technologies like Artificial Intelligence, big data, virtual reality, 3D printing and robotics.

Recent Government Initiatives in Data Science Education and Skilling

India has one of the most extensive education systems in the world. Though the spending in the education sector in the domestic circuit is not at par with that of developed nations like the USA, UK, Germany, Japan and China, the country is slowly catching up in the race. The Modi government, along with several government educational institutes, has taken significant steps to ensure tech skilling in artificial intelligence and big data. Many universities and organizations are expanding their data science courses and upskill offerings, in recent years. The government’s announcement to create data center parks throughout India, has been acknowledged as a much welcome move under Union Budget 2020. As data science continues to integrate entirely in the core business functions of the organizations, the demand for relevant skills and expertise will rise in the coming years.

Under the Big Data Initiative, the Indian government plans to carry out a gap analysis in terms of skills levels and policy framework. It also plans to develop a strategic road map and action plan clearly defining of roles of various stakeholders like government, industry, academia, industry associations and others with clear timelines and outcome for the next ten years, and carry out market landscape surveys to assess the future opportunities and demand for skill levels in same period. For this the Indian government is trying to understand the present status of the industry in terms of market size, different players providing services across sectors/ functions, opportunities, SWOT of industry, policy framework (if any), present skill levels available etc.

In the meantime, several government institutes have launched numerous educational and skilling programs for its student mass. Even the state government of Karnataka, Maharashtra, Telangana, etc, have come forward to enhance the big data talent pool in their respective states.

In April, Indian Institute of Technology (IIT), Kanpur started free statistics courses which included regression analysis, sampling theory, design of experiment and analysis of variance, linear regression analysis and forecasting, introduction to R software, among others. Later, IIT Roorkee teamed up with TSW, the Executive Education Division of Times Professional Learning to launch a PG Certificate Program in Data Science and Machine Learning. Around October, October, IIT Jodhpur launched a BTech program in artificial intelligence and data science from the academic session 2020-21.

IIT Madras also took an initiative in Data science skilling by offering affordable courses on Data Science through their platform ‘PadhAI.’ This five-month online self-paced course on Foundations of Data Science, ranges from beginners’ level to the mathematical and programming skills required for a Data Scientist.

Even IISc (Indian Institute of Science), in collaboration with TalentSprint, has plans to start a ten-month Advanced Program in Computational Data Science, with first cohort commencing in Jan 2021. The Computational Data Science course also comprises of a case study on the analysis of how the industry today uses computational data science in real-world use cases. The broader curriculum of the course includes the mathematics of data science, neural networks, machine learning, data engineering and business analytics.

Moreover, IIT Kanpur is offering two free online courses on Data Science, on National Program on Technology Enhanced Learning (NPTEL) online learning platform in India.

International Institute of Information Technology- Hyderabad collaborated with Great Learning to offer an 8-month extended postgraduate certificate in software engineering for data science. The program aims at helping learners to acquire skills in software and data science, especially technology leaders and managers who are looking to lead data science and AI implementations in the organization.

This year, the Central Board of Secondary Education (CBSE) announced the integration of Artificial Intelligence (AI) in the high school curriculum for the current academic year (2020 – 2021). Targeting approximately 200 schools for Grade XI and XII, across the states Delhi-NCR, Karnataka, Tamil Nadu, Orissa, Kerala, West Bengal, Andhra Pradesh, Telangana, Maharashtra, Madhya Pradesh, Uttar Pradesh, Rajasthan and Punjab, this curriculum has been developed in collaboration with IBM and will be a part of CBSE’s Social Empowerment through Work Education and Action (SEWA) program. The course will focus on teaching the basics and application of AI, and help with skilling in computational thinking, data fluency, critical thinking.

In July, the Additional Skill Acquisition Program of the Higher Education Department in Kerala introduced an AI Course of 776-hours, aiming to create skilled professionals who can fill the demand in areas of data science, artificial intelligence and Machine Learning.

In November, Telangana Academy for Skill and Knowledge (TASK) and Telangana State Council of Higher Education (TSCHE) partnered with Microsoft and NASSCOM to provide skilling 30,000 youths from Telangana state in Artificial Intelligence under ‘March to Million’ initiative. With an objective to enhance students’ employability by acquiring the skills required in the industry, the AI Classroom Series course introduces students to the concepts of Artificial Intelligence, Machine Learning, and Data Science.

Such initiatives are crucial as investing in long-term skills and capabilities can asset the development of existing human resources. It will also bring talent momentum while grooming the untapped potential of students and boosting big data maturity curve in India.

Published at Sat, 10 Apr 2021 07:43:13 +0000