Date: Thursday, May 30, 2024

Work Undertaken Summary

Risks

Time Spent

0.75hr EMA video 0.5hr time tracking 0.75hr TMA02 feedback 2hr communicating with tutor

Questions for Tutor

Next work planned

Raw Notes

EMA the final phase notes

Maybe I can have a documented handover meeting with KJ where we discuss the outcomes and what went well?

TMA02 Feedback

Page numbering - To try and find an automated way to keep the page numbers in sync.
Create a glossary of terms.
Include a diagram of what is in a context.
In the literature section, include ‘relatable’ e.g. ELI5 examples
Getting more people involved for greater validity?
1. Could have an interview with both a developer that is familiar with the area who can vouch for the accuracy, and with another developer who was not worked on the system to see how useful it would be with on boarding.
2. Hard to see how these conversations will be more than just “is it useful?” “yep” “do you understand this area now”, “yep”. Is that sufficient?
You requested to show some of the data transformations regarding the StoryTransformer section, currently I show the database schema (as a diagram) and an example output. Is this not sufficient or would you prefer seeing the database in another format, like the example in a series of tables and then the resulting output?

Regarding can I provide any interviews to back up the negatives of the current situation, I can’t really. The information is coming from me, because I am the person that is making the decisions / steering the ship / the person that is held accountable for what is ultimately delivered on the client project that my university project is about. I am the person I would interview. Do I need to make this clearer in the report that I am able to speak authoritatively on the direction, development, support and onboarding of the RecoWeb system?

Regarding getting 3rd parties involved in validating the success of the project I’ve got some thoughts but I’m really doubtful on their usefulness. I could have an interview with both a developer that is familiar with the area who can vouch for the output’s accuracy, and with another developer who was not worked on the system to see how useful it would be with on boarding. However I find it hard to see how these conversations will be more than just “is it useful?” “yep” “do you understand this area now”, “yep”. Is that sufficient / would this exercise be of any value?

What section have I chosen and why

I have chosen the Leads section as it represents a “medium difficulty case”. It involves functionality split over several screens and is split over 30 stories across 2 years of development and tweaks. This makes it relatively quick to iterate on with about 30 minutes end to end iteration. Therefore it strikes the right balance between proving the functionality of the system while being small enough to easily debug, understand and iterate. Additionally it is small enough to include all the input text in the submission.

Next Steps

Email tutor with latest progress and ask about the external validation.

Tutor Email

Hi Amanda,

The project is progressing well, in fact I’ve got some exciting news. I have completed a prototype for the StoryWeaver component which has allowed for my first end to end test. After some initial troubles with models getting stuck in infinite repeating sequences I’ve been able to get the system to produce a description of a single section about 9000 characters long which is genuinely useful! There is of course lots more development and refinement to go but the scary big question around feasibility is now resolved!

I’ve reviewed the feedback from TMA02 and I had a couple of questions for alterations I could make.

Regarding getting 3rd parties involved in validating the success of the project I’ve got some thoughts but I’m doubtful on their usefulness. I could have an interview with both a developer that is familiar with the area who can vouch for the output’s accuracy, and with another developer who was not worked on the system to see how useful it would be with on boarding. However I find it hard to see how these conversations will be more than just “is it useful?” “yep” “do you understand this area now”, “yep”. Is that sufficient / would this exercise be of any value?

Regarding what section I have chosen, I have chosen the ‘Leads’ section as it represents a “medium difficulty case”. It involves functionality split over several screens and is split over 30 stories across a 2 year timeline. This makes it relatively quick to iterate on with about 30 minutes end to end iteration. Therefore it strikes the right balance between proving the functionality of the system while being small enough to easily debug, understand and iterate. Additionally it is small enough to include all the input text in the submission. I will include this in my report.

You requested to show some of the data transformations regarding the StoryTransformer section, currently I show the database schema (as a diagram) and an example output. Is this not sufficient or would you prefer seeing the database in another format, like the example in a series of tables and then the resulting output? Or am I misunderstanding and you want a greater description of the steps that I am taking in the middle?

My new logbook entries, there are no developments or questions in these that haven’t been provided directly in this email. E.g. no reason other than curiosity to dive into them :) So just for reference here are my logbook entries in case you’re interested:

Thank you for your help and I look forward to hearing from you.

Best Regards, Luke

🪴 Quartz 4.0

Explorer

2024-05-30 - TMA02 Feedback