Useful Resources for Effective Set-up of Advanced Analytics, Data Science, Artificial Intelligence and Machine Learning Projects

Last updated 22-July-2020.

Just started consolidating curated, useful resources for this topic on this page, from my huge knowledge base. I am happy to share here list of resources I have compiled so far. Hope you will find some of these resources useful for your projects. I shall be adding more resources to this list in the coming days. Stay tuned for further updates.

Project Setup

• Setting up Virtual Environment for development activities
  • A Guide to Python’s Virtual Environments: What they are, how to use them, and how they really work
  • The Definitive Guide to Conda Environments: How to manage environments with conda for Python & R
  • …
• Container Platforms: Containers are a standardized unit of software that allows developers to isolate their app from its environment, solving the “it works on my machine” headache.
  • Docker: Docker simplifies and accelerates development workflow, while giving developers the freedom to innovate with their choice of tools, application stacks, and deployment environments for each project.
  • …
• Version Control
  • Git: Git is a distributed version-control system for tracking changes in source code during software development. It is designed for coordinating work among programmers, but it can be used to track changes in any set of files. Its goals include speed, data integrity, and support for distributed, non-linear workflows.
• Platform-as-a-service (PaaS) – Frees the application developer from the mundane tasks of installing and maintaining the hardware and software platforms on which the application runs. In the PaaS model, a service provider offers a fully managed platform on which applications can run.
  • …
• Skills, Project Roles and Responsibilities
  • …
• Managing Stake-holder expectations
  • …
• Planning and Tracking
  • …

TBD…

Successfully delivering Artificial Intelligence and Machine Learning (AI-ML) Projects

Coming soon…

• How to plan for AI-ML projects?
• How to setup AI-ML projects?
• How to do Risk management for AI-ML projects?
• How to ensure that AI-ML projects will be delivered successfully?

Operating Models – Digital vs Traditional

Coming soon…

• Power of Digital Operating Models – digital (AI and Analytics) at core, data-driven, agility, scale, continuous growth, innovations, network effect, ability to both generalize and specialize, …
• Traditional Operating Models – siloed, inflexible, complex, growth only up to a point then diminishing, specialization, …

Continuing high-quality education through online learning during and post pandemic

(originally written on 17-May-2020)

Authors: Suresh Dhamapurkar, Shilpa Dhamapurkar

Keywords: Educational Technology, Online learning, eLearning, MOOC, Pandemic, COVID-19, Instructional Technology, Instructional Delivery

Due to many advancements in technology and innovations in the Educational Technology ecosystem; education for all, anytime, anywhere has the potential to be a new norm. Advancements in compute, storage and networking technologies now enable education to be imparted online, at scale and yet maintaining high-quality if done correctly.

There are many reputed global institutions such as Massachusetts Institute of Technology (MIT, USA), who have been successfully conducting many high-quality online education programs since many years now, either through free Massive Open Online Courses (MOOCs) or Professional paid programs. Recently some of these institutions have even started professional Masters programs in online mode. Lot of R&D is going on globally, including in India in the emerging Educational Technology area.

Our Recommendations:

• Subject Matter Experts (SME) and Stakeholders:
  • Each college, institution should engage with Online learning (e-Learning) SMEs.
  • Ensure SMEs are hands on, they understand pain-areas of different stakeholders such as Students, Teachers, College administration, Parents, etc.
• Lessons learnt:
  • Learn from the experiences of reputed global institutions such as MIT, USA and many others who have been successfully running such programs since quite some time now.
• Tools:
  • Ensure online courses are designed and delivered in such a way that remote students are fully engaged and motivated.
  • Make use of good Learning Management Systems (LMS), Platforms and other tools in the ecosystem.
  • There are many technology options now available, facilitating students to interact with fellow students, teaching staff, extended online community staff as well as to participate in the discussions on certain topics of interest, asking questions, responding to other’s questions, etc.
• Assessments:
  • Imparting high-quality education requires significant efforts for designing assessment framework to test and assess students’ conceptual understanding and application of concepts.
  • Instructors can consider designing a good combination of quick exercises, practice problems, case studies, graded exams, peer-reviewed assessments, adaptive assessments, etc.
  • Choosing the right set of tools is important for conducting such tests. Assessments if designed correctly would help students gradually reach towards achieving his learning goals for the course.
  • As of now, technology and tools are not yet matured enough for conducting proctored exams remotely, at scale seamlessly. Until good quality and affordable solutions are available in the market, consider blended assessments models.
    • Critical exams and practicals which cannot be done effectively online, can be conducted physically following social distancing and healthcare norms, may be in a number of batches.
• Performance Analytics and Personalized Learning experience
  • Evaluate online learning tools for critical features such as measuring and proactively tracking administration, teacher and learner performance.
  • Recent advancements in Big Data, Advanced Analytics, Artificial Intelligence and Machine Learning technologies enable personalized learning for students.
• Training
  • It is essential that administration, teaching staff and students are trained on using tools efficiently and effectively.
• Diversity and Inclusive:
  • Creating a diverse and inclusive learning environment is critical for building a strong, growing nation.
  • Facilitate needy students, who can not afford to have good networking connectivity and the right environment at home to take courses remotely.
• Competitive Advantage
  • Through superior innovation in terms of high quality personalized learning at scale, helping each student achieve learning goals tuned his/her potential; online learning has the tremendous scope to positively impact lives of many students from any corner of the world. It offers massive competitive advantage to the educational institutions and organizations.

(image source: Photo by Julia M Cameron from Pexels)

Mobile Social Networking

Access by anyone to everyone from anywhere (using a mobile device) will radically change the user experience. “Building smarter systems to give a better experience to users” is the challenge. Atos sees mobile social networking as a fundamental application for social communication.

A mobile device increases the value of social networks as they also become mobile. Due to continuous connectivity, social networking with a mobile device is more symbiotic than with a desktop computer. Atos believes that mobility is already offering social networks new opportunities, like real-time communication on the move, location-based features, and the sharing of personal information (contact lists, calendars) via a mobile device anytime, anywhere if users wish to do so.

This introduces a host of new business possibilities, such as becoming a mobile social networking platform provider or an application provider. Business models (between application provider and end user or between application provider and platform provider) could include per transaction, sponsorship, advertisements, onetime payments, micropayments, payment per subscription or even per click, per visit or per-use.

Source:

Based on whitepaper from Atos’ Scientific Community describing how social networking using a mobile device is different from that using a non-mobile device; the benefits and use cases; the ability to share and collaborate in real time with the real world; and the challenges and business opportunities in this area for consumers, service/platform providers, and application providers.

Authors: Suresh Dhamapurkar (Chief editor), Santi Ristol, Mahesh Chandrasekar

12 ways you can achieve efficient and predictable service delivery

1. Take care of your team – address their concerns, recognise, motivate, mentor, groom, empower, encourage team working
2. Define & agree clear KPIs, RACI matrix with clear responsibilities and accountability for all the stakeholders
3. Practice Lean – Increase value added services, minimise non value added tasks, eliminate wastes
4. Use optimum span of control for managers & team leaders and optimum resource pyramid
5. Follow knowledge management best practices
6. Use standard tools and processes
7. Reuse -frameworks, approach, solution, components, etc.
8. Ensure team mix with right competencies are deployed
9. Ensure right level of governance
10. Define and implement quality gates
11. Automate, semi automate repetitive tasks, reports where feasible
12. Ensure tight control over startup activities for new project/phase or service until steady state reached

Why both productivity and efficiency measures are required for Application Management (AM) & Application Development (AD) bench-marking and tracking output?

Productivity is defined as the ratio of total output produced vs total input supplied.
Efficiency is defined as the ratio of total output produced vs actual efforts spent to produce this output.

In the software application management world output can be measured as number of incidents/queries/bugs resolved, number of standard changes delivered, number of batch jobs scheduled or monitored, number of small/medium/large enhancements delivered etc. Application development work output can be measured as number of Function Points delivered, number of objects (screens, reports, batch scripts, etc.) delivered, total LOCs delivered, total LOCs migrated, etc. Input is typically person efforts (hrs, mandays, etc.).

Efficiency of an individual or a team to resolve say an incident would typically involve actual efforts spent to resolve an incident. Whereas productivity of an individual may also involve waiting time for clarification resolution and other overheads which are not directly adding any value to incident resolution.

So efficiency is required to track real efforts required to produce a unit of output and productivity is required to track what additional efforts are required for other non value added activities or inefficiencies due to waiting for work, infrastructure, no clean order, etc. Highly efficient is the one who produces best output by consuming minimum possible resources.

Organisations need to measure both productivity and efficiency to create a baseline, to monitor performance against baseline and have plan in place for continuous improvement of these important KPIs. They need to focus on getting productivity of individuals/groups in their organisation closer to the efficiency by doing right things to work towards the purpose.