#save me (pls)
1 messages · Page 1 of 1 (latest)
and then tailor your projects to that specific role
like full stack
or backend or front end
\etc
ye
okay lets see
education is fine i dont think fundamentals of cs matters but thats nitpick
skills should go below exp and projects
at the very least below experience
but experiences --> projects --> skills flows much better
deleted clubs and affiliations noone cares unless a) very specific experience for a very specific roles (" i lead and created the core programming club at [t100 school] and ended up getting sponsored by nvidia to create a hackathon" type of shit at which point that go in ur exp or at the very least projects)
elixir — Today at 12:47 PM
trying using xyz format instead of having multiple bullets that address different parts of it
you've accomplished (X) + the qualitative results (Y) + the skills or experience you utilized to achieve the outcome (Z).
lets look at ur exp bullets
what does development and engineering support mean
so i can rewrite accurately
yeah essentially
so the company was a game dev company?
and this was game dev internship that ur tagging as SWE?
no
it’s an educational company that teaches high school / middle school students cybersecurity concepts through game dev
application dev is more aws stuff
i’m still trying to figure it out
Created and implemented diverse environments in [company]'s cybersecurity game-based educational platform utilizing Unity, C#, and [whatever else , id definitely say smth like procgen methodologies and basically say u used them for inspo] to enrich user experience
rough idea of xyz
that talks abt a) what u did b) its impact c) how u did it
dont play down what u did ever
play it up always
using the word support in this context is just throwing it in a box that says "i helped them with this product and it had a lot of users"
if we want to put stats into this
Created and implemented diverse environments in [company]'s cybersecurity game-based educational platform utilizing Unity, C#, and [whatever else , id definitely say smth like procgen methodologies and basically say u used them for inspo] to enrich the experience of 500+ users daily.
and obv if thats too long shorten it down somehow rewrite it w/e but thats the format and idea
makes sense?
kk lets look at the others too
developed and designed physical evidence related challenges w unity webgl
probably archaic now
with ur first point basically covering it
guessing theyre the same thing ?
yeah
that means u can expand on what u did even more
now u have another free bullet
ur bullet 3 is gone now too
i think
yeah probably
but u can prob turn it into a bullet about hygeine in unity and industry standards since thats actually genuinely important
if u have enough to write
or
you can talk more abt ur AWS stuff now
instead of dedicating 3 bullets to what can be condensed into 1
sure but the thing is the aws work starts like next week
so i can’t talk about it in depth
its fine talk abt what you will be doing and bullshit a bit
jsut make sure to edit and update a little
if u can go too indepth its not a big deal but u can probably find some bs that will make sense in the lense of what the company does
knowing its just placeholder as u apply during this cycle
thats up to you
if you wanna expand more or not
but now you have more room for more depth on what youve actually done
be sure to fill it and show me when ur done if u want more feedback on it
i probably will
just so it’s not only like a game dev experience
projects arent bad but ill take a better look in asec
think im seeing same issues
liek this can all be 1 bullet
was it for CFG
btw
or does capital one have other hackathons
nah c1 summit
kk
created an advanced budgeting website for Capital One's summit hackathon, efficiently accessing user data through custom API implementation and a thoughtfully designed React and HTML based frontend
smth like that
x: created advanced budgetting . ..
y: efficiently accessing user data
z: custom api, react html frontend
and then u can also have a separate data point maybe talking about how you parsed through the data you access with the api
okay
how many bps should i go for per project?
3 bullet points that are all like this
or this
will be fine
and some stuff that you talk about wont fall into xyz
like if youre talking about what you learned
sounds good
dont xyz that itll sound montonous and robotic
but 3 strong bullets is generally fine
go thru all ur projects, try and fix this stuff and come up with more to write, and bring it back to me and we'll workshop it some more
works for u @vale hound ?
yessir thanks big man
and remember exp + projects above skills
bet
just ping me
i might take a bit to respond im working on my portfolio proj rn
but i will asap
bet
question @half elbow , for the autonomous robot project, i really only worked on it for a little while and I didn't really do anything meaningful per se, but I do plan on returning to the project this fall in like a few weeks. is it worth keeping it?
yeah
unless u want to do more projects hten u can prob make a better one on your own
like computer vision x arduino
but if ur doing projects the first one to go should be game
im guessing u used p5play
if u did use p5play anyone that recognizes p5 will know that p5play handles collisions / physics for you ( i think ) and itll actually just be taking up space
if u manually did collisions thats better
regardless it’s like the most basic project
yeah
i can prob cook something better in a few days
so keep the automation stuff
unless you plan to rework projects
then make ur own arduino shit thats better and u actually do work on or update it as u do work in the club
and do smth better than the js game
kinda depends on direction
if u want to stay in crud do crud
if u want to do low level stuff, another robotcs / core / wtv thing
don’t really want to tbh
if u wanna do game dev, a new game
i’ll prob try some low level stuff
also wanted to make some ml stuff
like really basic ml
u can link ml arduino and low level
if u do
a computer vision project hoisted on an ardunio hardware that has the robot react to its environment based on what it sees
let me think a bit
ur right it is basically 2nd project
u shouldnt replace teh 2nd one just make smth new for first
2nd is fine as long as u can elaborate more since realistically the amnt of work u did isnt quantifiable and they cant fact check
im gna go exercise ill b back in an hour and we can talk more abt potential projects back then
if possible try to have the reworded bullets for exp, proj1 , proj2 done by then
no rush tho
sounds good
gimme a sec
i fixed the technical skills and reworded bp's for experience and proj #1, still figuring out bps for project #2
also swapped out fundamentals of cs with intro to computer organization
oops forgot to tag @half elbow
add another bullet point to the c1 summit (figure smth out ik u can)
bullet points for club projcect need to be fixed but we can go thru those together
game bullets can be fixed but alos since u have al ot of white space the solution may be to first build another project then work on replacing game
also
dev tools and frameworks/libraries
can be combined into 1 line
just labeled tools
and remove vsc / visual studios
ides dont matter
here's mine
there r more tools i could add
but i rly dont need to
tkinter/nltk is fluff
need to take that out
but u get hte picture
yeah just like the important ones
yep
and u have some good stuff to add
jira , aws
r good ones
this doesnt look exactly like jaeks resume
its not right
or is it?
looks a bit different
its jakes but i changed a few things
i forgot exactly
maybe a different font
i also changed the color of the borders
yeah kk
just make sure to utilize that white space
to sell yourself
every inch of this should be selling yourself and making what u did seem insane
so make another project
i thought on some let me type it out
moving more in direction of machine learning :
arduino handwriting recognition , object detection and categorization , voice recognition and categorization based on traits like pitch resonance etc
the thing abt ml is its recognition + reaction, so many things will be detection and response projects
yeah
if u wanna focus on low level stuff or robotics stuff (this is why having a direction is important)
handwriting and voice recognition would be cool
thing is, i don't think im experienced enough to know what I like yet
kk definitely ml then
then we can also step out of the scope of arduino based
if thats smth youd pref
yeah im cool with that
also i could prob replace the game with a basic image manipulator class project i did
if thats any better
then like :
image recognition and sorting from medical database that looks for specific signs of xyz to categorize them into folders
game engine (Basic project , dont do this , just an example )
nlp using transformers /llm to tap into generative conversational AI
detecting anomalies in normal networking from the lense of cybersec
they both suck
but the game is atleast more unique
yeah
u can put it on for now though
before u apply to more
since u have nothing else and you have space
okay
yeah
library based projects r mid generally
but
everyone starts there
and ur 2026 grad
u have a year to turn urself into god
altho
def want to start asap
im 26 grad too and im still stressing asf
wagmi
this first one looks really cool tbh
u should do it then
its a good project too
and the execution could be as simple as like
also im working on like a big website based proj soon
so with that and another project it should be decent
different file folders labelled my name of condition (ie like 3 folders, 1 named GIST , one namde liposarcoma, one name leiomyosarcoma) and u just sort the images into folder based on what the llm recognizes , if it doesnt recognize any create a new folder for the name of what it find --> put it in there
if u rly wanna do more u can also have it highlight or circle vi image manipulation / pixel manipulation or wtv and then replace the old file then store it
okay, now the question is how do i get started
i have like 0 ml experience
it just seems like something interesting to me
gpt will give u a list of like
12 steps idk
ill tell u what my query brings back
btu the cool thing is
u can literally just ask it to elaborate on the step
and based on the prompt you can also be like
Show me all the frameworks used
i need to start using chatgpt more
yeah its broken
iut will hold ur hand
through the entire project
its code gen sucks
but
it will tell u everything u need to be able to learn it yourself
thats what i need ngl
thats sick
yeah
and u need common sense cus gpt is stupid as fuck but it has access to more information quicker than you do
so use it and vet it w ur brain and ur good
yeah its very confidently wrong sometimes
i used it to help me w one of my projects
(smarthome dashboard in python django nodejs that utilizes node-red to control smart devices in your house given the alexa api key its connected to)
full stack app otok me like 5 days
its broken though cus i suck at front end but its good enough to talk about
and the cool thing is like
thats a dope project
gpt holds your hand but you genuinely need to learn what youre doing to make it
so ur not even gonna be in a place where like
damn i made this but i dont know how and i learned nothing and i will fail at any AIML internship
here's gpts response to the first query btw
thats hella detailed though
give me a list of models i could use for this
ChatGPT
Certainly, here is a list of deep learning models that you can consider for your medical image recognition and sorting project focused on detecting specific signs of condition XYZ:
Convolutional Neural Networks (CNNs):
CNNs are the most common choice for image classification tasks. You can start with various CNN architectures, including:
VGGNet (VGG16, VGG19)
ResNet (ResNet-50, ResNet-101)
Inception (InceptionV3, InceptionResNetV2)
MobileNet (MobileNetV2)
DenseNet (DenseNet-121, DenseNet-169)
Xception:
Xception is an architecture that builds upon the Inception architecture and has shown strong performance on various image classification tasks.
EfficientNet:
EfficientNet is a family of models that have achieved state-of-the-art performance while being more computationally efficient compared to earlier architectures.
ResNeXt:
ResNeXt is an extension of the ResNet architecture, designed to improve accuracy by increasing model depth through a cardinality parameter.
SqueezeNet:
SqueezeNet is a lightweight CNN architecture that achieves good performance while having a small model size.
WideResNet:
WideResNet is an alternative to deeper networks, emphasizing wider network structures instead of increased depth.
DenseNet:
DenseNet is known for its dense connections between layers, allowing for feature reuse and efficient training.
NASNet:
Neural Architecture Search (NAS) models like NASNet use automated methods to find the optimal network architecture for a specific task.
3D CNNs:
If working with 3D medical images (e.g., CT scans, MRI), consider 3D CNN architectures like 3D-ResNet or VoxResNet.
it will do the entire thing for you
just ask it to expand on steps
"expand on step 3, what models can i use?"
"how do i access this model [model that you picked]"
"how do i set up this model"
"im running into htis error during setup how do i fix?"
"back to the original steps, expand on step 4"
etc
yeah
okay so to do list
- fix bullet points for every project so far
- decide the project ur gonna make
- condense frameworks libraries tool into just tools, remove fluff
- add image manipulator and write bullets
show me
gna go shower
all makes sense ?
its not a website or anything btw
legit just code that adds sepia filter, shrinks, and fades images
okay
it and the game are gonna get replaced at some point anyways
yeah
kk gna go shower work on this ping when done if im not busy ill respond otherwise we'll workshop tomorrow
pce
sounds good thanks again
Will look after food
add another bullet to software developer intern
more projects , there's so much room so we have lots to work with
def can fit atleast one project in there
u ml project will go next
like put it in before ive done it?
have smth displayable even if rudimentary or even if just the backend algorithm by abt a week if u have time
if u dont realistically have the ability to show anything b the time an interview could come of this resume (~1 week) the ndont put it on
i def can
should i deploy it on a website or just host it locally
yeah
but website is clickable by them
ill focus on the model first and then work on the website
yeah that one is the hardest for me to fix because of lack of experience
bet
collaborated in team setting to create an autonomous r2-d2 robot (x , what youve accomplished) that was able to navigate campus (make some shit up about why you did it maybe campus safety or maybe testing to see if it can map out areas idk thats all you) using a combination of C++ and Arduino to handle firmware and hardware interaction.
one bullet point and u got the whole thing
didnt leave anything out
the elss bullet points you have for each (that contain everything) the more projects you can make
yeah
why does it say r2d2 robot
gotta spam projects
is that the name of the robot
hi speemch
that looks a little funny at first glance
its supposed to be r2-d2 from star wars
like designed after it
yeah
i think its open cv
but i havent started yet
we were working on making it remote control first
oh makes sense
and then making it autonomous
autonomous is hard lol
yeah its a big team
could lowkey get a paper out of it if u get it working. right
oh makes sense
with like a team of 6
yeah i was trying to make filler for purepixel
1-2 bullets is fine
ill condense rn
gives u more room for more projects
experience needs a lot of bullets
projects just need descriptiveness
yeah that makes sense
i just dont want any whitespace
but that just means i need more projects/exp
yep
whitespae is good because you have room to put more
whitespace with 5 proj is bad though
cus how
but
we'll look when we get there
lmao
developed an image manipulation software that allowed the user to efficiently apply a variety of effects to images of multiple filetypes utilizing OpenCV, NumPy, Pillow
smth lke that
thats rough
rly rough
but u can gpt it if its giving u aids
just make sure its human enough and accurate
also the opencv and pillow stuff i added today
cus it was hella rough without them
code was light though
thats fine its good optimization and looks better
yessir
hmm
unfortunately desertdrift isnt much u can do
cus like there's not even much to write in 1 bullet
yeah its pretty bad
its fine itll get better just replace it asap
i can add a small unity game ive done but i feel like thats basically the same
and i dont really want two diff unity exp on res
order of priority during free time should be
- barebones ml proj we talked abt or wtv u decide
- replace desertdrift with ANYTHING (barebones, but pref something that gives direction for what u want [ai ? ml ? low level ? compiler ? etc ] )
- flesh out ml proj
- flesh out replacement to desertdrift
i would say to use the website ur building as ur replacement to desertdrift but
2 full stack apps neither of which is insanely complex on back end is kinda pointless
its kinda like checking the "i can read htmlcss documentation and make it interact with a backend" twice
its an education website
if u rly wanted u could put the c1 engineering summit in exp
and the full stack proj in its place
🤷♀️
thats ur choice
tho
hmmm
esp since u have more to say for it than just a generic project
yeah it was a lot of stuff
(working in team , databases, mock api, parsing, team work [?] did u work in a team ? , arduino, watever)
yeah
if ur gonna move it to experience though u absolutely need to capitalize on all the space u gain for it
yeah i agree
however, im probably going to prioritize ml project
fullstack is probably going to be a weekend thing
yeah whatever works for u
i have like a week before i go back to school
im technically 23 or something but i move in the 18th i think
a little more lke two weeks
good
lots of time
fuckin loads of time
knock out the ml in a week
10+ hrs a day to do it no reason not to
is that possible?
yeah i def can put that much time towards it
especially do the degree you want to do it (barebones)
research will take u longer than coding it will
what algo to use, how to sort, what database to train it on, etc
yeah
but u can make something demonstratable in that time
yeah once im bnack to school im not gonna have as much free time
this is like the best time
alright tldr
1/ reduce projects to as few bullet points as possible without losing info (reduce fluff, condense )
2/ put in description shit / project for ml , grind it
3/ create full stack app (education or wtv) and move c1 summit to experience (f u want to move it to exp, thats ur call)
4. show me
works ?
and maybe try to add another bullet to ur exp but if its not coming dont force it
yeah
education project won't be done for a while though
haven't started yet
like i have a figma for it
thats about it
imma add c1 summit to exp, just checked one of my connection's resume and they have it on exp too
also is software dev intern better than software dev for res purposes?
yeah id say so
but i think thats negligible
unsure on that
intern is most likely better though based on my gut
kk imma get off but once u do this @ me , if u need help on any of it @ me
alright thanks
can y’all review my resume lol
im not much help but elixir is the goat
sure i can look but ur cracked asf dont know how much i can help
dm it to me tho or post
@half elbow i found a good dataset for skin cancer detection, but that lowkey might be a bit advanced for a starter proj
i could do plant disease detection which is far simpler
that works too
bet
@half elbow lmk what you think
maybe try formatting the relevant coursework bullet point in the same way the experience/projects bullet points are formatted
fixed it
@half elbow whenever u get a chance ^
better
now lets optimize more
sec ill draw it out
if you've already moved onto a second line, you may as well use all the space to give a more accurate description of the project, or describe more about it
so we can either a) probably shorten some points to one line
or b) fill out the rest of the lines for ones that cant beshortened
yup yup
i can definitely expand on projects #1, #2 once i finish them
just gotta grind those out so i can speak more about them
good
for desertdrift
remove the word "accurate"
in the 2nd line
and it'll become 1 line
all the words after hte comma of the last point --> enhancing player immersion
if thats still too much , user interface --> U.I.
or
interface*
try and do that kinda thing fro each line and then where you find yourself still being on 2 lines, fill out the rest of teh 2nd line
yeah
although if we try hard enough we can condense all 3 into 1 bullet probably
but
its a placeholder
yeah perfect
less sapce is better
all the bullets on my res are 1
except for a few
how many projects do you have?
my projects seem a bit odd because im a game dev student not swe so i have augment reality, platformer
about to add a custom gameboy advanced emulator
8 projects total 🔥
pefect
legit talking about this saved me in an interview today lol
talking abt the research ?
he asked me about python experience and i started talking about the image manipulator and plant disease stuff
LOOOL
thats perfect bro
good shit
thats what we like to see
add python into the techs at the top of it tho
appreciate you
🙏
im in trouble for soph programs like explore though
imma have to take off my experience
ill prob keep c1 stuff though its not really exp
but that means i gotta add more to cover up whitespace
i gotta capitalize
probably
wagmi pal
but non-urms get explore all the time
i hope so