If you have a vertical SaaS product focused on a particular industry or a SaaS product focused on a particular functional area you are probably thinking about how Artificial Intelligence (AI) can be used within your product. There are a variety of product strategy and technical issues that will need to be addressed and of course you should only incorporate AI if that is the best way to address your customer’s business issues.
Once the above issues have been analyzed and a decision has been made to include AI in your product there are implications for successful customer on-boarding and the professional services required. This is especially true for applications where the AI application is based on data that is unique per customer.
The reasons to have professional services in you SaaS business such as ongoing customer satisfaction, customer retention and up selling are all even more applicable in an application with AI. In fact, it is hard to imagine any type of B to B SaaS application without professional services.
If you have an AI platform product used in other applications then there are a different set of issues that need to be addressed. This post addresses B to B SaaS applications that need minimal software development for the solution.
Professional Services for SaaS implementations are comprised of both the people who do the work and the process used and AI has implications for both.
There is a level of incremental expertise and understanding that will be required by all professional services staff. The level of understanding and the specifics will depend not only on the product but also on the role. The areas will vary depending on the circumstances but would typically include:
- Understanding of how the AI technology is used in the product
- Basic knowledge of appropriate training data-sets
- Understanding of any additional legal and ethical issues associated with the product as a result of AI
- Understanding of data management tools for data cleansing, organization and movement
Typical professional services roles can include solution and technical architects, developers, project managers and analysts. Additional expertise in how to architect and develop solutions and project management of AI implementations will be needed. For architects and developers this is primarily technical skills.
For project managers the projects will likely tend to be more complex due to the following factors:
- More focus on data preparation and data management
- Less predictability of project length due to
- Unpredictability of the time needed for data preparation
- Unpredictability of how long and how many iterations will be required for AI learning
- Projects will likely be more iterative with the number of iterations being less precise due to the data related tasks
- Change management and education for the customer including end users, technical staff and management
A least one additional type of role is likely to be needed in professional services, the role of some level of data scientist. This role will need to be involved in all aspects of the project including the early scoping and most planning phases. Ideally there will not need to be a data scientist dedicated to each project or you will have problems scaling the business.
There will be a significant amount of technical, product and process training that will need to be implemented for professional services staff so that the staff can successfully evolve into the changing roles.
Process differences will include changes in project structure and additional focus on client management parts of the process.
- Project structure
- More data preparation and management tasks
- Likely will be more of an Agile project management approach due to iterations involving learning from data sets and measurement of a successful implementation.
- Project length likely will be less predictable
- Data-sets for AI learning and solution implementation
- Attention to contracts & who owns what data and what data can be used in various circumstances
- Vendor vs Client responsibilities for data related challenges. The obvious answer is that the client is responsible but they may not have the knowledge and expertise.
- Setting expectations
- Length of project
- Agreement on what defines solution success
- Change Management
- Client training
- Time to deal with potential client resistance to AI
- Business Issues
- Less predictable project size implies T&M contracts
- Less predictable project length & size stresses scheduling processes
Although there are substantial changes to people and process, the underlying administrative tools are unlikely to be substantially affected. However, the changes may stress the implementation and associated processes of using the tools. There will also likely need to be changes in what technical tools are used.
The biggest overall changes will be an increased emphasis on training and education, an increased emphasis on client interaction skills and changes to the actual solution implementation process and projects.
2 thoughts on “On-boarding for SaaS Artificial Intelligence Applications”