Architecting Student Success: Integrated Software for the Community College

A Q&A with Bob Bramucci and Jim Gaston

As the mission of community colleges moves from access, to access plus outcomes, campus leaders look to software that supports student success. Bob Bramucci, Vice Chancellor, Technology and Learning Services and Jim Gaston, Director of IT-Academic Systems explain South Orange County Community College District's strategies for its development of "Sherpa." Known primarily for its success as a recommendation engine, Sherpa is actually an entire "sandbox" of integrated software that provides new tools for student success.

Mary Grush: In designing Sherpa, SOCCCD developers chose a somewhat different approach than you'll find for many other systems that push out information or make recommendations to students. Could you talk a little about your development choices?

Bob Bramucci: In our initial development we chose a human expert-based approach that is somewhat reminiscent of the early days of artificial intelligence work. What this does is to codify the decision rules of human experts and automate them so as to vastly increase the reach of the human expert. We certainly value--indeed we treasure--the expertise that our staff have, but there is a limit to the number of advisees an individual can handle. Counselors and advisors are very scarce resources, so we were looking for a way that we could magnify the reach of our experts, by developing an expert system.

We can and will extend and augment the system, though, and we are now also working on a predictive or learning analytics project that is still in its early stages. But at this time probably the biggest differentiator between Sherpa and any other recommendation engine I've seen, is that it depends very heavily upon the wisdom of human experts, trying to automate and extend that.

Jim Gaston: Another significant choice that we made was a conscious decision early on to try to approach our development more as architecture than as a final, discrete product. We didn't want to presuppose how Sherpa would be used and wanted to have flexibility for the future. Several things have come about that bear out the wisdom of that decision. Our use of and needs for Sherpa have already taken some turns we may or may not have anticipated, but all of these things can be more easily accommodated because of this open architecture--SOA [service oriented architecture]--approach.

For example, in the past couple years Sherpa has been used very heavily for making general announcements. The actual targeting of specific services, tasks, or information to specific students is a more recent activity that we are doing mostly in some very controlled pilots at this point. But with SOA, other systems can be integrated with Sherpa as needed for these purposes as we move more towards targeted, or personalized, student services.

In the state of California there is an enormous push right now to focus on academic planning. Our existing MAP ["My Academic Plan"] tool is a very helpful planning tool, though in the past it was not used to push out recommendations to students. Now, we are in the process of an integration between Sherpa and MAP that takes advantage of our SOA approach so that we can proactively push individualized academic planning recommendations out to students. There has been a wide discussion across the state, among many community colleges that are not sure how they are going to address the academic planning mandate. We were able to answer this question and start working on the integration and development in a relatively short period of time.

And of course there are many other functions and tools included in our ongoing development of Sherpa. Because of our open architecture, and the ability to integrate Sherpa with other resources, we are able to take, for example, the same recommendations generated from the MAP integration I just mentioned and insert them in other areas--for example, right in the midst of registration when students are selecting their classes. One of the key objectives of our state's Student Success Task Force is not just that the student creates a plan, but that they follow through on the plan--having these recommendations tied in with the registration process is bound to be helpful.

Beyond e-mail and text nudges, we also have the ability to add tasks to students' To Do lists, which has been very well received by students. The students can complete many of the tasks immediately online through services they can access via their phones or a Web browser on their computers. And we have another proposed project that we are seeking funding for right now, to create a student success dashboard that will gather several factors that are critical to a student's success and put that information right in front of them.

Bramucci: It might be useful to note that though the recommendation engine in Sherpa has probably gotten most of the public attention, that's not the way we think of Sherpa at SOCCCD. We think of it as a student success sandbox, where different software engines play together nicely using a set of rules. We actually started the Sherpa project years ago when we built a search engine for our classes--that was even before the emergence of recommendation engines. So now and into the future, Sherpa will actually include a mix of software geared for student success.

Jim already mentioned one of the most important examples: that we are integrating our existing academic planning engine, MAP, with Sherpa--adding it to the sandbox. And the new predictive analytics work we are starting will add a math layer to the human expertise layer of Sherpa. So we see Sherpa as very extensible. And as Jim pointed out, Sherpa is not a single product--it's a student success software sandbox where several software engines can collaborate towards the goal of student success.

When we originally chose our metaphor for Sherpa, it was the Sherpas of Nepal, who selflessly help other people reach their dreams--the Sherpas are content to stay in the background. Getting back to Jim's point about architecture, while I talk a lot about Sherpa with other IT professionals and in interviews, and even though everything in our portal is being served out by Sherpa, the students still never see the name 'Sherpa'. It's in the background, just like the Sherpas of Nepal.

Grush: It sounds like there are so many possibilities of what those software engines might be. How do you guide the development of Sherpa?

Gaston: We have a design team that is comprised of administrators, managers, faculty, staff, and most importantly students, who have guided the priorities of Sherpa from the beginning. We didn't want to open the floodgates, certainly. And anything that we do, we want to be able to measure. We want data that will reflect the degree to which we are making a difference for our students.

The whole purpose of Sherpa is to intervene just at the point when it really means something to the student--right when we can help them make good choices and act on them. So the design team decided to focus on a set of recommendations from the California Student Success Task Force--recommendations that relate to very specific student actions that are measurable. We can measure our success if you will, at helping our students be successful.

Grush: Will you be making Sherpa available to other institutions?

Gaston: We knew early on with Sherpa that we were on to something that is going to greatly benefit our own 43,000 students every semester. But we do have even larger goals than that. We would like to assist students at other institutions beyond our district, and again our SOA approach helps make that a realistic potential. Sherpa is tightly integrated with our own, homegrown Student Information System, but it is loosely coupled--which means that the integration points are very well defined. And there are no assumptions within Sherpa that the data that it's getting is necessarily coming from our SIS. All of that means that Sherpa could be integrated with systems at another institution, whether those systems might be PeopleSoft, Datatel, Banner, or theoretically any other SIS. Our hope is that Sherpa will eventually be shared as a cloud-based, hosted service.

Grush: Then would you intend to share Sherpa with community colleges in particular?

Bramucci: We are currently in discussions with the California Community Colleges Chancellor's Office. We would like to make both Sherpa and MAP available system wide--free within our system throughout the state. But we are still interested in productizing outside of the California Community Colleges. Not that I wouldn't want to open source it; it's just that with the current economic climate development money is getting very hard to come by, and we still have lots of good projects we want to fund. So, we are seeking a revenue stream so that we can continue to develop new student success software tools that our community colleges urgently need.

While I see that the software we have created is not at all limited to community colleges, I do see a particular relevance and value for community colleges. I would liken what we are seeing in community colleges statewide, to trying to turn an ocean liner around. Ever since the Master Plan, community colleges have had a role to play, and that role was in a word, access. We joke that we are very selective: "We only accept the top 100 percent of those who apply." That is very different from most baccalaureate institutions. And for decades, our mission has been access, access, access. I believe that we have succeeded remarkably with that. But now, for the first time in half a century, this system is being reconfigured--now aiming not solely at access, but for a mix of student success while preserving access. That's a greatly changed mission. And the California Community College System has 112 colleges in 72 districts, and 2.3 million students--making it the largest single system of higher education in the nation. That ocean liner is not going to stop and turn on a dime. But I think everyone recognizes that keeping the access while increasing positive outcomes and working on student success is where we are headed.

Only about one third to half of America's community college students reach a positive outcome. That's at the heart of what we are trying to address.

[Editor's note: SOCCCD's strategies for Sherpa will be featured in a breakout session presented by Bob Bramucci and Jim Gaston at Campus Technology Forum 2013 in San Diego, April 29-May 1.]

Featured

  • person signing a bill at a desk with a faint glow around the document. A tablet and laptop are subtly visible in the background, with soft colors and minimal digital elements

    California Governor Signs AI Content Safeguards into Law

    California Governor Gavin Newsom has officially signed off on a series of landmark artificial intelligence bills, signaling the state’s latest efforts to regulate the burgeoning technology, particularly in response to the misuse of sexually explicit deepfakes. The legislation is aimed at mitigating the risks posed by AI-generated content, as concerns grow over the technology's potential to manipulate images, videos, and voices in ways that could cause significant harm.

  • close-up illustration of a hand signing a legislative document

    California Passes AI Safety Legislation, Awaits Governor's Signature

    California lawmakers have overwhelmingly approved a bill that would impose new restrictions on AI technologies, potentially setting a national precedent for regulating the rapidly evolving field. The legislation, known as S.B. 1047, now heads to Governor Gavin Newsom's desk. He has until the end of September to decide whether to sign it into law.

  • illustration of a VPN network with interconnected nodes and lines forming a minimalist network structure

    Report: Increasing Number of Vulnerabilities in OpenVPN

    OpenVPN, a popular open source virtual private network (VPN) system integrated into millions of routers, firmware, PCs, mobile devices and other smart devices, is leaving users open to a growing list of threats, according to a new report from Microsoft.

  • interconnected cubes and circles arranged in a grid-like structure

    Hugging Face Gradio 5 Offers AI-Powered App Creation and Enhanced Security

    Hugging Face has released version 5 of its Gradio open source platform for building machine learning (ML) applications. The update introduces a suite of features focused on expanding access to AI, including a novel AI-powered app creation tool, enhanced web development capabilities, and bolstered security measures.