Reclaiming SWAYAM

Today’s news article on the SWAYAM MOOCs and open-ness by Anil Sasi of the Indian Express raises some very important questions about the future of MOOCs in this country.

The facts of the matter are as follows. A proprietary rather than open source approach has been adopted because open source seems not be open after all. Choosing EdX, for example, they believe compromises intellectual property and requires a big fee to be paid to MIT (even after EdX, at the behest of IIT Mumbai and MHRD gave over the full source code and support to India in 2013 and assured that all IP will remain with India). Secondly, it seems they believe that open source systems do not have the depth of being able to handle enterprise grade learning environments. Third, this is the conclusion of expert committees of the government after in-depth deliberations, I assume, with a wide range of industry, technical and MOOC experts. Fourthly, the RFP itself built by PwC and the government, the basis of the INR 38 cr project award to Microsoft, is in itself plagiarized and deficient.

This defies logic. A really large part of the world runs on open source. The open source movement has shown that enterprise grade, mission critical applications can be made to work with community support. Total cost of development ownership is lower with use of open source. And open source, by definition, fosters collaboration and innovation.

At the risk of repetition, instead of manufacturing large systems, the government should invest in building API and making integration possible between systems. They should fund edTech startups to build MOOC based learning environments. They should enable an open architecture, not just in technological terms, but also in terms of an open architecture of participation.

How would that work?

On the technology front, let us assume we are API focused. Then we must openly build the following API sets (and more):

1. User API – API that allows users of different types and institutions to be managed, for different stakeholders and their roles
2. Identity API – that allows users to be uniquely and securely identified through the course of their life, with probable integrations with other systems like Aadhar
3. Curriculum API – API that enables metadata and classification systems for content and pedagogy, that brings Corporate, VET, School and Higher education taxonomies together
4. Assessment API – API that enables taking online assessments of different types, enables proctoring controls, provides secure test-taking and great analytics
5. Certifications/Badging API – that allows certification/degree providers to create online badges and certificates that can be awarded; secure lifelong eportfolios and linked certificate depositories
6. Authoring API – that allows quick and easy authoring, review and collaboration
7. Content Delivery API – API that allows video streaming (live and VOD), CDN-grade access, shared folders and cloud distribution
8. Network API – that enables social discovery, network and group formations, sharing and amplification and social profile aggregation; building both social and learning graphs
9. Services API – that enables tutors to connect to students, mentors and coaches to their mentees, institutions to parents and so on, and provide services such as fee payments, digital and offline educational content, tutoring, adaptivity, virtual classrooms and so on.
10. Andragogy/Heutagogy/Pedagogy API – that enables different techniques to teaching-learning to be used as desired by teachers and students, e.g. blended models or SPOCs.
11. Learning Analytics API – that provides new ways of deciphering engagement, learning and interaction.
12. Language API – that enables multi-lingual content and internationalization

(Remember that technology and all this talk about API is merely the greasing in the wheel. The real work is in exploring new paradigms of teaching and learning, especially online and blended. And this does not mean building online courses and calling them MOOCs.)

These API sets (and others I may have missed) would need to be supported by a strong developer program, funds allocated for several incubation initiatives with participation from private funds, R&D labs, education programs to build engineers and architects of future learning environments and many more. important aspects known to us from the experimentation & learning of the open community in discovering what works at scale.

Now imagine a time when these API are available (in fact a large number already are available in the open domain, they just need to be contextualized in some cases) for use by indigenous developers. They are not starting from scratch. They are not restricted by a monolithic RFP or scope. They are not constrained to be this one very large proprietary solution (although some may want to build such systems on top of the open stack, which is just fine). If things go well, a number of people will focus on developing alternative solutions to pieces of the puzzle, while others will integrate them into solutions that can be used in different contexts. No one size fits all.

This will give a boost to indigenous development, which at the current time is laboriously trying to build each component. It will bring about that strategic 10x inflection in edTech in India enabling thousands of providers, who are operating mostly in isolation, to get a framework around their efforts and build for scale. Strategic funding for R&D will help us achieve breakthrough innovations in teaching and learning at all scales. Private sector funding of edTech will find a purpose.

This is what the government should do. And only a government can achieve this at strategic scale, tying up all the piece of the supply and demand chains, particularly in a system so dominated by public education.

Read Full Post »

The corruption of the SWAYAM RFP

It seems the SWAYAM RFP dated 21st November, 2015 is actually inspired from previous RFPs made for other contexts. You have to only compare the SWAYAM RFP with two earlier RFPs:

1. National Career Services Portal RFP dated  13th August 2014
2. A JNU RFP on eLearning Development dated 5th February 2015

 

To give a sense of the malaise, here are indicative architecture diagrams from the NCSP and SWAYAM RFPs. Try and spot the differences.

You don’t have to be an expert to recognize copy/paste. A simple Google search is enough to lay bare the blatant plagiarism. The consultants for this RFP in turn may have been inspired by others across ministries and their appointed consultants.

But there are deeper issues here.

Firstly, the very respected Pricewaterhouse Coopers (PwC) has been hired as the consultant to draft the RFP, select the vendor and monitor the implementation. It is possible they acted in similar capacities for one or more RFP consultations. To find PwC indulging in cheap copy and paste goes against the very reason they were selected.

Second. PWC is not doing this free of cost. The entire exercise is expected to cost MHRD about 30 lakhs with NICSI also getting a slice. With the efficiency that comes from copy and paste, one would think the effort would be far below proposed.

Third. It may also be okay, to copy and paste certain generic specifications. But would you propose the same technical architecture for two very different contexts? Worse, would you ignore the advances in technology over the past two (or more) years and be content with copying older ideas?

Fourth. Even while doing a copy and paste job, would you at least take care not to repeat earlier mistakes made by the earlier authors. The mere act of a copy and paste indicates an intellectual vacuum. When done improperly, it indicates the complete absence of intellect and intention. Take for example the following diagram (look at the circled phrase). Laughably, see how Sentiment Analytics, the subject of much excitement in the recent past has now become Sentimental Analytics!

Fifth. It is not very clear if PWC was the perpetrator of the earlier RFP or other similar ones in the past. And whether they were paid similar astronomical sums for their obvious consulting expertise to copy and paste.

Sixth. While the government can take a hands off position and blame PWC for these acts of omission, there is no way be not held accountable for their choice of consultant, for their inadequate review process and for other errors of their act of commission. The MHRD must explain how this travesty has occurred with full internal and vendor accountability. It is scary that we are going to invest so much public money and effort in an initiative which seems so flawed from the word go.

Seventh. I have not yet even talked about the actual content of the RFP itself. It is so obviously incompetent that I can only sigh with frustration at this phenomenal display of MOOC and technology expertise. And I am not talking about the Microservices vs. SoA kind of higher level technological debates either – just very simple things that I daresay most MOOC technology people would be happy to point out are missing, erroneous or irrelevant. It would be superb to place the panel of experts who edited or wrote the original version of the RFPs in a public debate, asking them to substantiate their proposals.

More galling than any other thing is the obviously brazen attitude that anything they do will pass public scrutiny. There is perhaps a babu-consultant-OEM racket in here which I hope someone takes the pain to uncover. Perhaps they genuinely believe we are idiots who will not really care.

I sure hope we are not.

Read Full Post »