Incoming MS, looking for both internships and new grad roles

[iron torrent]

mf censored his school with "top public CS school" im done

[twilit mulch]

@eager shuttle Purdue?

[twilit mulch]

GT?

[twilit mulch]

curious lul

[midnight coral]

Michigan

UMD

UW Madison

[dry robin]

ok i think this is p good

but there are a couple things i woulds

change imo

one i would personally remove

the bolding on the

resume bullets

cuz its p distracting imo

also i'd pro specify

which nlp model

u used for

project

do u know which models u used or like what type of networks they were?

also im assuming u did a bit of feature engineering?

from what i've

seen from ur first bullet

so i'd prob go into more details about that if possible

also personally i would prob add some metrics like accuracy for this project imo

is this paid?

i put deep learning resaerch from my school on it cuz i get paid for it

imo

idt training time should be the biggest metric

imo

cuz there is sm factors that come into play that can make it non-deterministic

like im assuming this a binary classification problem right?

like personally i would include accuracy on the valdiation set

[livid coral]

what is the diff between the 2 resume's

[dry robin]

this is like an example of binary clasification problem

i have on my resue

like usually when ur validating a model

u often split ur data into a training and validation

sets

well this was more of personal project i did for fun, i found a dataset from nih

so i figured why not make my own model for fun so i can get better at signal processing

gotcha

https://www.deeplearningbook.org

this is very good souce imo

if u want to get somewhat of a primer

i think it's fine to keep but

there a lot of things

i would personally change

at least for descript

hm i would personally be more descriptive if possible

if u havent done too much yet

i would prob jsut get rid of it

if ur curious on how to format it, this is i how formatted a research internship i did after my freshman yr summer

hm are u trying to get more into data science?

oh personally, i would prob leave that project off ur resume if possible until u do more advanced project cuz imo i think it's too trivial and overdone imo iyk what i mean i.e. same as calculator made in react

idt there is specific type of project that is better than another as long as it's what somewhat demanding
if ur aiming for quant dev, imo a low latency project or hardware related project would be good
for big n or general swe any generic full stack/front end project should be fine imo
for trading, maybe something dealing with numerical methods or more hardcore ml architectures

it's basically some software project that is designed to reduce the delay between communication time between large amounts of data or dealing with hardware
ig one example would be a distributed system where u have some unified memory that is connected to each processor and n processors having their own local memory, often times when ur building algos for these types of systems u want to minimize the communication time of sending data between these processors as it can lead to large overhead cost and designing efficient network topologies to do so or algos can be very challenging task

[dry robin]

just ooc did u finish ur bs?

cuz u titled ur thing as incoming ms

hm gotcha

ok

so one thing i prob would change is i would prob just hypertuned

model

instead used spaCy and panbdas

cuz that doesn't really make sense iyk what i mean?

[woeful cloak]

I don't like resulting in significant cost and time savings or improving efficiency and overall software reliability... you should demonstrate using metrics for these

otherwise it's a nice resume 👌🏼

[meager ginkgo]

I would ideally remove your coursework - you've graduated with your BS and have enough experience/projects under your belt that you don't need it.

You have fairly strong bullets overall, you can resolve some issues with "texture" by bolding tech keywords (common) or metrics (less common) or both - it'll make it easier to parse

My recommendations for reductions/edits:

• "Spearheading adoption of contract testing using Pact..."

• "Collaborating with senior engineers to develop Developing a smoke testing service that enables XXX engineers to easily validate critical functionality~~ of their applications, improving...~~"

• "Trained and fine-tuned pre-trained Developed/Trained/Fine-tuned (it doesn't matter) NLP models..."

• Here's an opportunity for using "AI" as a buzzword, since you used NLP models before: "Created an AI chatbot using Python that used NLP models to automatically..."

• If you're not going for full stack roles, you can cut down the tech used for bullet 1 in role 3 (also no TypeScript in languages?) to make it less bloated.

  • "Developed an interactive web dashboards web application to visualize customer orders for inventory management using React, TypeScript, and MongoDB, leading to an 18% reduction in overstock costs"

• "Designed and implemented a cloud-native system..." (also add metrics for speeding onboarding process)

• "Built a CI/CD pipeline that go hand-in-hand..."

  • This bullet overall needs to be rewritten, a lot of it doesn't say much, and you need to make the purpose clearer and how it saved 20 min per deployment

• "~~Designed multiple game levels that ~~incorporated environmental obstacles..."

  • latter part of this can be simplified, so this bullet hits one line, up to you

I'd also be adding the rest of your tech skillset into your skills (new section for Tools) such that there's no confusion on what you know.

Minor nit, but I'd also remove the line divider under your name, just because it's making your resume feel a little more packed without uninterrupted whitespace in the header

I'd also rethink the "Designed a cloud-native system" bullet because it's not 100% straightforward what it does (it's better if you meant automate instead of streamline, but otherwise, it's not clear)

The bullets for your insta clone are the weakest (strangely enough) in the entire resume, just because (1) it doesn't solve a unique problem space and isn't a unique solution (2) doesn't have users (3) uses SSG for what's supposed to be a dynamic app? and (4) testing quality and coverage is more important than "to enhance software quality" - consider updating (I have no immediate fixes for this)

It's fine, I strongly doubt you'd get rejected for it, but as a full stack specialist, I am puzzled by your approach the moment you brought up static site generation

It's probably the first thing I would phase out given your experience in other areas once you build other projects/gain more experience, and the second bullet can be cut if you feel space is tight

It's something I would probably prefer to discuss in an interview, like stuff you learned about SSG in general

That'll depend on the roles you apply for too - make sure your projects are tailored to it

You do have a strong resume overall, so I wouldn't be too concerned - you're just at the stage where you have to be more targeted with your applications and stuff, esp since you have pretty good work as a generalist

Sure, I just don't know how helpful I'll be since I'm not an academic lol

[pseudo garnet]

Looks good to me!

does the instagram clone have any users? I know it's a clone so probably not, but i'm curious

[dry robin]

looks fine

[pseudo garnet]

Coursework

[dry robin]

congrats

hm if u get return offer maybe

just take snowflake and graduate

but otherwise not bad idea

to extend offer

yeah for ft

also i wuldnt worry too much about loan

u can def pay off

w new grad job

hm usually pretty rare imo unless if its part time

[dry robin]

i think as long as top 20 cs if u have faang or big n u should be good, decent amount of state school ppl at my last summer intern just had only faang

quant with faang

nvidia/snowflake

def a notch above faang

either way still decent company