All team members present for the retrospective. More small group async in week, based on current work.
Frontend
Jessie
Bansharee
Backend
Kevin
Yizhen
ML
Kevin
Yizhen
Amittai
Recommendation Engine
Kevin
Amittai
Deployment
Colton
Kevin
What worked well
Sticking to small team meetings, continuing distributed workload.
Continuing practice of high documentation (recommendation engine)
Using Tuesday meetings for long term alignment
Using GPT4 to automate some tasks (assigning cross topic complexities and affinities -> what are relation between topics?
What didn't
FrontEnd
Using SwiftUI for Watch instead of UIKit - SwiftUI provides too little control over playback, complex swiping interactions require lower-level control
Multiple videos play concurrently on swipe, unstable playback experience
BackEnd
ML
Updated scraping to use API to produce video object
Recommendation Engine
The model we are trying to use is somewhat massive, so we had to figure out ways of making it less heavy computationally but still retaining its utility.
Deployment
Self-assessment on progress
Where are you in relation to progress towards product and milestones?
FrontEnd
Watch won't be fixed by Technigala, will require a couple days afterward to fully transition old Watch (written with SwiftUI) to new Watch (based on Instagram's IGLIstKit)
Have good sources and packages
Some endpoints still need to be linked to backend API
BackEnd
ML
Need to modify back end slightly to match ML
Recommendation Engine
We spent a lot of time working on vectorization of our videos in a way that we can query it for similarities, and building the surrounding infra for the recommendation engine.
Deployment
Give an estimate of how far towards your goals you are, do you think you're on track?
FrontEnd
Backend linking half done
Some pages/features need to be redone, including Watch
Behind, won't finish by Technigala but confident we'll finish before end of term
BackEnd
ML
Scraping near final, deploying on EC2
Recommendation Engine
We are confident that we should have it working end-to-end by the end of the week.
Deployment
Lay out each of the following weeks till end of term with brief goals for each
3 weeks left
Rec System / Webscraping (video and transcript) / Topic Alignment /
Algorithm Clipping / Backend Link to Rec /
Rec/ML/DB Link / Frontend Backend Link
Deployment
Keep working on Watch, link to Postman, link to backend
Sprint 7 Retrospective
All team members present for the retrospective. More small group async in week, based on current work.
Frontend
Backend
ML
Recommendation Engine
Deployment
What worked well
What didn't
FrontEnd
BackEnd
ML
Recommendation Engine
The model we are trying to use is somewhat massive, so we had to figure out ways of making it less heavy computationally but still retaining its utility.
Deployment
Self-assessment on progress
FrontEnd
IGLIstKit
)BackEnd
ML
Recommendation Engine
We spent a lot of time working on vectorization of our videos in a way that we can query it for similarities, and building the surrounding infra for the recommendation engine.
Deployment
FrontEnd
BackEnd
ML
Recommendation Engine
We are confident that we should have it working end-to-end by the end of the week.
Deployment