Skip to content

Week 2

Date: 23.02.2026 - 27.02.2026

Role: Dev in team Tracing

Activities

  • Tracing pipeline implementation

    • Action: Implements mulitples parts of the tracing pipeline defined in (Document file (PDF))
    • Artefacts: Class Tracer (permalink Github)

    • Justification: I worked more specifically on the following methods :

      - load_texture()
- load_model()
- palettize_texture()
- split_colors()
- compute_fill_slices()
- format_palette()

    • Skill related: Applying computer engineering skills to the field of data

  • CEO'meeting report
    • Action: Produce the meeting with client report
    • Artefacts:
    • Justification: Keeping trace of the content of a meeting with the CEO serve multiple purposes. First, it lay on paper commitment made from one of the parties during the meeting. This can for example, protect the team from a delivery that the client could refuse, if it's form was previously written as such in a report. Secondly, it serve also as self-reflexive tool to analyse points to refine for futur encounter.
    • Skill related: communicate clearly and effectively

Blocker & resolution

  • Problem: The meeting with the client was a disaster. Unpreparation and no responsible led to very few questions/decisions taken.
  • Solution: The client is not an engineer. As such, we need to vulgarize a maximum and help his decision with facts, prices and visualisations. We should go as straight to the point as possible, everyone's time is valuable.
  • Problem: A specific librairie : Pymesh was missing documentation and methods.
  • Solution: Look for alternatives online and ensure that documentation is well maintened and for example issues are responded in Git.

Self-reflection

What went well?

  • The tracing team successfully implemented the planned pipeline.
  • The daily meeting were shorter than last week, so we were more concise and on the point.

What could be improved?

  • Preparations for meeting with the client. Due to the lack of accomtability (no leader) nobody really prepared it.
  • Due to communication issues, some people's work during the week was worthless. Having team leader with overview will help

What did you learn?

  • that the client has very limited time/abilities to understand engineer language
  • that responsibility is a motor that should be imputed to members of the team
  • that testing different plausible solution require a well define test pipeline that can be automatized to cover large panel of cases
  • that a person with overview is needed to ensure avoiding blind double work inside a team

Proposal of interview questions

Q1: What project, work related to data science are you particularely proud of ?

A: A programm called BlueRook. It was a chess game, made with pycharm interface and in python. I did it in my last year of college. Although the quality of the code in the project is worthless, it gave me the orientation of the branch. I went on until the point of having an opponent playing by itself mode. There, I was already refering to RL as further implementation possible.

Q2: How do you view a successful meeting with the client ?

A: The team arrive with the needed decisions ready to be presented to the client and get clear choices from him. That is vulgarized, sourced alternatives, with clear decision making path. The meeting is then successfull if every needed questions/decisions have been answered/made by the client and all of it is recorded in a report.

Q3: In a team effort involving 8 personn like this project, is an horizontal leadership a good way to lead the team toward completion ?

A: No, in an horizontal leadership environement, simultaneously everyone and nobody is responsible for anything. This will certainly lead to unprepared situations. For example, if nobody is clearly responsible toward success, then the urge to work could be decreased. A system with a group leader and specific dev's team leader will ensure more accountability to the work and ensure a better flow.