Hiring for VIVO projects

Hiring for VIVO projects

Typical VIVO-related job postings

One big aspect of starting a VIVO project is to find the necessary staff to do the work, and the VIVO team receives many questions about the skills and level of effort required.  There is no one-size-fits-all answer, since the IT and application support environments vary so widely from large research universities and government agencies to small schools and non-profits.  Before developing job descriptions it may be helpful to factor in the range of tasks and decide where your current team most needs to be augmented.

VIVO surveys (written by Paul Albert at Weill Cornell Medical College) have included questions regarding the amount and division of labor that participating sites have needed.  It's telling how many different skills can be associated with a VIVO project, and also how much variation there is in the range of answers.

For the respondents, the average total for the 11 roles described is 1.95 FTE, with a maximum per responding site of 6.5 FTE (estimated at 4 FTE ongoing for a major research university) and a minimum of 0.2 FTE (from an institution in the exploratory stage).

It's also likely that some roles such as project champion, marketer, or even project coordinator may not be officially accounted for but added into the jobs of people already on staff.

The programmer/ETL specialist has the highest average commitment of effort at 0.5 FTE, which certainly makes sense, followed by the project manager/coordinator at 0.3 FTE.  IT/systems engineer at 0.19, content coordinator at 0.16, ontologist at 0.15, and marketer/outreach at 0.12 follow.

One frequent question is how many sites have specialists in each of these areas with suitable fractions of FTE available to work on VIVO?  More likely a small group of multi-talented staff will have to be assembled to cover these 11 roles plus any other locally-specific tasks.  How the tasks will break down depends on the individuals involved, but tends to settle into three principal areas.

Policy

Most VIVO projects need a leader who may not be the executive sponsor but who has access to a network of campus relationships and can translate big picture goals into achievable steps suitable for the timeline and resources available. Most often this person becomes the project manager to coordinate the work of a content specialist and a programmer/web developer.

Very often this person will do most of the outreach to build the case for VIVO and promote early prototypes or case studies.  Talent at writing and presentation is important; having the skills to produce appealing marketing materials would be wonderful but may often be brought in on an as-needed basis – look for a talented visual designer who can establish a clear branding that can be applied both to the VIVO application but also to related outreach materials.

Content

The content role requires a detail-oriented person who can cope with data that will almost inevitably be found to be in worse shape than anticipated. Ontology skills are useful if a site plans to introduce new types of data, but otherwise the focus can be more on data analysis, on building communication and trust to have access to all the right information from available sources, and on quality control as first steps expand into production quality, repeatable processes.  The content coordinator needs to work well with the available technical support but usually doesn't need to understand the nuts and bolts of scripting.  It's more important to have someone look carefully at content both before and after ingest to be sure information should be public and is correctly represented.

Technical

The IT support environment varies widely from institution to institution.  VIVO has not typically been a difficult application to install and support operationally in an environment where there's "devops" support and systems engineers are in close communication with application owners and understand their preferred technologies. As the data scales, however, there can be some more challenging configuration and resourcing issues to address in optimizing the use of processing, database, and memory.

When an IT support environment already exists, VIVO technicians typically focus more on data ingest, customization of branding, ways to re-use the data. In a small organization without a history of supporting enterprise web applications, the skill set of the VIVO technician may need to encompass cover aspects of server and database configuration outside the scope of typical web developers.

The primary technical challenges in building up a VIVO stem from acquiring, transforming, and updating data from institutional systems of record.  A good ETL (extract, transform, load) person will find the tasks familiar but may find the transition from working with row/column data tables or hierarchical XML structures to RDF and the Semantic Web to have a steep learning curve.  We recommend allowing time for a person to train in the basics of RDF, the principles of ontologies, the different mindset of a triple store, and the SPARQL query languages before trying to immerse themselves in complex data transformations. It's essential to know what the target is – what RDF conforming to the VIVO-ISF ontology should look like – before jumping into writing mapping tables or XSLT transformations. This is often best done by entering a small amount of sample data into VIVO interactively and exporting the RDF to look at what has been produced. Once the task and targets are clear, the programming is not especially hard, so look for someone with clear thinking and a willingness to learn.

Do you need a Java programmer? Note in the chart above the distinction in roles and required FTE between the IT/systems engineer and the programmer/ETL specialist – while the VIVO application is written in Java, the skills required to implement a VIVO focus more on scripting and data ingest, especially after initial installation. And while a Java programmer may be more familiar with writing and deploying applications similar to VIVO, it's familiarity with the application stack (principally Apache, Tomcat, and MySQL) more than Java programming that will be required.  Some implementing sites have extended the VIVO software itself for extended functionality by modifying the Java source code (notably the Deep Carbon Observatory project at RPI), but most may never have to change a line of Java.

Sample job descriptions

• Technical job descriptions
• Content job descriptions
• Policy job descriptions, including outreach