CNM Tech Board 2024-01-11

From CNM Wiki
Revision as of 17:28, 11 January 2024 by Gary (talk | contribs)
Jump to: navigation, search

To stretch out

The following four strategies are proposed to discuss:

Market-first

Let's more aggressively take into consideration potentials of our product to bring customer's attention, interest, desire to try, or, eventually, buy. If the potential of any product is low, we should consider its retirement.

IA-focus

In the last weeks of 2023, we discovered AI as the missing piece of our puzzle. Initially, I thought that AI is something really-really complicated and out of our scope. Basically, AI shall become the core of what we do. I plan to utilize both commercially-available models like ChatGPT and create our self-hosted models as well. Our students will play with the models and utilize them to generate content and get basic knowledge. We shall use more and more of AI; particularly, we will divide expert assistance in two parts -- (a) AI-driven expert systems to enhance our work and (b) expert contractors. No contractor shall be hired until we exhaust AI aid.

More apps

Because of better infrastructure, we should also add more pieces of software. Any new piece of software increases our capacity to serve our upcoming interns. And, I am really positive that a few of those apps will be AI-powered; I surely expect our own chatbot, possibly, within a couple of months.

Better collaboration

We shall deploy some software to boost our collaboration. Out of open source, three are highly regarded. (a) Focalboard (teamwork management) is simpler than Taiga. It has meeting support features and can show its content to the public, which is really cool. Moreover, it is developed by the same team that develops Mattermost, a well-known messenger. (b) Taiga.io features an AI-powered teaching assistant that is designed to help school students learn software engineering in a fun and interactive way. Taiga also can be integrated with our GitLab. Finally, (c) GitLab also has its project management features. Based on my research, GitLab has a suite of AI-powered features such as code suggestions, code explanations, vulnerability summary, and chat. I don't know which of those are available in their self-hosted version. I see many proprietary tools such as Taskade, Asana, Wrike, and so on that get AI-powered. No open source tool does so yet, but we know that open-source are normally 6-7 months behind their proprietary brothers, so we should expect enhancements. Once again, we need to assess all of those three and start using them. As soon as we can, we will also augment it with AI.

Gary prompted ChatGPT-4 whether Focalboard accommodates board meetings. This is the answer,

Yes, Focalboard can be used to accommodate meetings. You can create custom boards for different projects, use automation to save time, invite team members to collaborate, use tags and labels to categorize tasks, and use the calendar view to schedule tasks. Focalboard can also be used to create recurring meeting agendas, like team meetings and 1:1s. Here is a sample meeting agenda template that you can use:
To Discuss | Revisit Later | Done / Archived
Meeting participants can queue new items under “To Discuss”. As you go through them, most are moved to “Done”, but a few need follow-up. You can move those to the “Revisit Later” column. The owner then moves it back to “To Discuss” for a future meeting.

To monitor

Tech board

Our current meetings are good as prototypes, but we need to refresh the ways we plan, conduct/record/stream, and document/close them, as well as act on them. True -- board meetings are complicated phenomena. Besides conducting they meetings, they involve (a) video recording, (b) streaming, (c) minute-taking, (d) video publishing, (e) scripting, (f) agenda preparation, (g) task filing, (h) Internet connectivity, (i) microphone-provisioning, (j) well-running Jitsi or Nextcloud -- whatever we choose, and, of course, (k) marketing. The last video didn't impress me with regards to the quality of my audio and video. I am committed to doing whatever it takes to improve those, but I need my to-do list. Another issue, which I have been stressing for years, is that our recording is not conducted in a central place. Right now, the quality of my video, for instance, depends on my connectivity multiplied by Natalia's connectivity. If we locate both Jitsi and recording at one place, we will improve the overall quality. Back to marketing, the boards shall generate attention and, possibly, interest in our project. From my perspective, it must be AI and software rich and well-done. If we cannot make it well-done, we should retire it as CNMCyber events.

Community board

As of now, I view our community board as an opportunity to engage with people of our core interest such as the tech lady in the school that we are interested to partner with. In terms of marketing, it is created not to attract new customers, but to attract new partners and team members.

Hands-on training

On the upcoming Saturday, Sergey will introduce me to their 90-120 minutes training. As of now, Gary uses this page -- Basic AI training to self-learn.

School-student products

Gary proposes that aggressive schedule -- (a) one week for last minute consultations, (b) one week for negotiations with marketing agencies in Eldoret and creating a legal framework, (c) a couple of weeks for scheduling the first cohort -- possibly, anything from 12 and up to 50 students -- enrolled for free, (d) each week, we should start a new cohort, (e) after the first successful week of practicing, start charging parents, initially, nominal fees, and increase as the program matures.

Employability

Gary updated his resume with AI features -- Gary Ihar -- and encourages his teammates to do the same :)