2024-01-08 Chris Exit Interview

Owned by Daniel Puratich
Last updated: 2024-05-14 by R D
4 min read

  • Meeting

    • @Christopher Chung

    • @Daniel Puratich

    • 9pm pst loool (both of our first days of work)

  • Timeline

    • joined team Jan 2022

    • became lead after two terms later (July 2022)

    • stepping down starting Jan 2024

  • Onboarding Experience

    • he was very active so knew what was going on

    • wasnt too much handoff between him and Dhruv

    • were a couple calls, but it wasnt a leadership guide, more of a task sync

    • give a lot of advice on what the role of lead is and how to fill that role

      • tried to give that to derek

  • Director Influence on Projects

    • current director stack is very diverse and passionate (so good)

    • nature of EFS is small role in competition so not a lot of director direction

      • hopefully will start to change as EFS does more development on COTS

      • currently it’s kind of subpar

    • Thoughts on the ZeroPilot program and development

      • EFS should put more resources on COTS so it contributes to WARG better, was neglected as the team made a more COTS shift

      • before the team was very focussed on custom flight controller firmware

    • Milestone 1

      • was flying manually in fixed wing

      • this was what WARG had before we restarted the team

    • time for zp to replace ardupilot and cots stackup is a few years

    • biggest stuff for past few years on ZP was that ZP kept restarting, kept rewriting for a long time

    • takes a lot of dedicated members & leads (leads for architecture) to get stuff done

    • big diff between wARG and other teams (like rocketry) is the firmware is at a large scale and not individual

      • at rocketry each board kinda does one thing and sits on CAN bus

        • one person can do a single thing by themselves

        • each person kinda gets their own repo

      • for us we have one chip doing everything, lots of devs doing for one chip

        • here the scale is much larger

    • not a full time thing so people develop and leave for a week or two, and it rlly kills stuff for review time

  • Team Direction

    • past and future projects

    • Hopefully COTS projects are more small groupy instead of ZP which is a very large group

    • Chris like COTS being at forefront with ZP in parallel

    • lots of members for it

    • more individual boards from EE

      • CAN boards are a good example of this!

      • good for people to get out of ZP and more into a board of their own

    • PMs & Growth

      • able to find PMs for all major projects

        • all zp managers

        • sim

        • hil

      • picking active people is better then picking knowledgable people

        • willing to learn is most important

      • hardest part is retention, espeacilly long term retention

        • knowledg transfer is crucial

      • happy with subteam direction

        • hardy & derek are clutch (strong leads)

      • lots of good newer members

        • need to retain them

      • ofc lots of people join and drift away

  • Leads & AEAC Meetings improvements

    • meeting structure

    • leads meeting last term was short and sweet

    • Daniel had column and chart before hand, so people could look before and after, people can focus for sections they care about

    • saving time is important

    • didnt talk much during leads, a lot of it is admin stuff so lots of just listening

    • leads last term was prolly run best way possible

    • went to just a few AEAC syncs,

      • didn’t feel like needed to go

      • didnt get a lot of info out of it

      • didnt find them as important

      • just bc EFS isn’t involved in it

        • more cots projects will have it shift

      • this kind of speaks to Chris’s involvement with the team

        • chris never went to a comp

        • comp was just an extra thing that occurred

        • cares more about mentorship and development and architecture then the comp stuff

        • changing teams direction will change people’s minds

          • chris was into ZP just bc team was into ZP

          • say it how it is in the info sessions, be real with people, attract people who are interested in what we’re doing, if it’s more COTS we’ll get people for it

          • chris doesnt know a lot of cots stuff, more fundamentals, so this step down is a solid shift

  • Sysint team, do we need one

    • a sysint team would be

      • really helpful

      • good

      • somewhat neccessary

    • no matter what we want thsi team it’s just nobody wants to do it

    • the only people that do it are people who are super invested in team and passionate

      • Aidan fell into this bucket, but when he got busy it fizzled out

    • example:

      • on efs there’s a lotta infrastructure work

      • linting, github actions

      • everyone knows we should have it

      • nobody was willing to do it until derek

      • gotta find the right person to get those benefits, just tough to find this person

      • it’s technical, but also tedious

        • slightly unsatisfying

        • lots of chasing people down

        • not a lot of concrete accomplishments

    • potential plan fwd

      • wait a bit

      • find someone who is interested in the job description

      • can’t rlly trick someone into it, must tell them what it is

      • mostly chasing deadlines, architecting the system, not a lot of cold hard engineering/designing

      • Aidan was perfect for that role

  • directors shouldnt be scared bout changing things on EFS

    • it’s reasonable if directors make the hard decision to table ZP we can

    • dont want them to be circling on ZP ins

    • talking bout ZP

      • ZP is a long feedback loop, project is like at least a few years dee

        • LOS forced architecture, very complicated

        • We could go even simpler (we dont wanna restart)

          • but if chris could do it all again, just start with bare metal loop

          • a lot of stuff doing rn is good practice (future proofing) but complex

            • example: free rtos (not rlly required for a lot of simple stuff we tryna do)

          • more interation would be good

            • maintain functionality

      • the mech frame gets feedback at most every year for comp

      • hoping COTS projects have smaller feedback loops, better for learning, maybe better long

      • will play it by ear going fwd

  • thats it