Skip to content

The job of Solution Architect

Solution Architects are consultants. The IT Consultant is defined as "knowledgeable professional who helps businesses develop, integrate, and maximize the value of IT systems". I do not want to re-invent all the extended information about IT consultant job, but for me there are two main group of activities:

  1. Manage day to day tasks like any other white collar workers do: meetings with customer, issues management, project management, dependencies management, management reporting, internal administration tasks.
  2. Knowledge management: As Stephen Covey in "the 7 habits of highly effective people" wrote: we need to sharpen the saw, to be a better person. This is where the value of a consultant resides. SAs share his knowledge to get customer’s trust, so customers feel comfortable to invest in your technology and in your company because they feel the company's technical human resources will help them to remove roadblocks. Therefore Solution Architect needs to continuously work on his skill set for IT, Cloud and different technology to support customer's requests and build trust.I think as of now a SA should spend around from 40% to 60% on knowledge acquisition and maintenance.

In 2023 I think the role of SA needs to change dramatically bad adding generative AI as part of the job.

To simplify the management of the knowledge acquisition and measurement process and get better results per consultant, genAI will help dramatically as it will offers an integrated and standard way to share knowledge. It will help scaling SA business too by improving the quality of the responses and increasing the number of accounts per SA.

We should be able to get rid of those useless certifications, that do no proof anything in term of consultant value: it just demonstrate that the person was able to memorize a lot of questions and answers to succeed to an exam at a given point of time. 80% of this knowledge is lost after 2 months.

After 2 weeks we tend to remember 10% of we read, 20% of what we hear, 30% of what we see, 50% of what we hear and see, 70% of what we say, and 90% of what we say and do. Those classical management numbers will stay valid for a long time, and demonstrate we should enforce show and tell as a model for knowledge acquisition.

The big idea will be to build on top of existing LLM the following corpus for SA:

  • Service product documentation
  • Blog, forum, re:post, Q&A, FAQ, articles
  • Labs, code sample
  • Architecture patterns
  • Slack channel content

The consultant will leverage this chat to get answers to most of the customer requests. The skill will then evolve as understanding customer problems and questions, assess what is the context of those questions, query the bot and select the best answer, then validate the answer with hands-on work. As to apply show and tell, we should focus on getting SAs the most efficient tools possible to be able to demonstrate and build PoC very quickly and the same way cross SAs.