I'm speaking from my past experience as a hiring manager at a start up with outlier standards for performance and trajectory in software engineering and machine learning. I estimate I've screened tens of thousands of resumes and interviewed at least a thousand people in my career.
First and most important: your internships and work experience, and what you accomplished during those jobs. They should tell a story of increasing and accelerating personal growth, learning, challenge and passion. If you can share personal or class projects, even better.
After your experiences, your degrees will be considered based on the number of years each typically requires, with early graduation and multiple majors being notable.
1. PhD, if you have one. A STEM PhD was particularly helpful for ML/Data
science positions, but not required.
2. BS/BA (3-4 year degree)
3. MS/MEng (1-2 year degree)
Put another way, if you don't have a PhD, the MS/MEng program is a tiebreaker compared to your experience and undergrad credentials.
International students get a raw deal. The online masters will barely help you get a job or launch a career in the US. US universities appear to offer the chance to work for major US companies with a notable university (such as Georgia Tech) on your resume, only to feed their graduates into our broken immigration and work authorization system, H1-B indentured servitude and no replies from the countless companies that have an unspoken higher bar for those needing sponsorship.
To round out a few other contexts HN readers might experience:
If you are an international considering an on-campus MS/MEng, US universities are charging full price while giving you a credential of limited value and utility. Apply the same comments above but at a much higher price than GA Tech’s OMSCS.
If you are completing/just completed a less notable undergrad degree, paying for a masters program at an elite CS school (like GA Tech) is usually a bad deal. If it not a requirement for the positions you seek, it won't help your career chances much.
If you have an undergrad degree and your employer will pay/cover your MS/MEng at night/personal time (and that is your passion), awesome and go for it! It will be a lot of work and lost sleep to get everything out of the experience, but a lifelong investment in your growth and experience.
If you are completing/just completed a notable undergrad degree (tier-1, internationally recognized program), you don't need the masters. Feel free to get one for your learning, sense of self and building research connections while you ponder getting a PhD. The hiring and salary benefit will be very small--you are already the candidate every company wants to meet. If you decide to get a PhD, that will open some new doors but take 5+ years to get there.
At my previous company, we made it our forte and team passion to get authorization for employees--given a global pool of candidates and a hiring bar to match. I'm really proud of our effort here given the broken and unfair system. Sadly, many companies do not share this value or cannot justify the time, effort and expense, or cannot scale such a program to a larger number of employees across a less selective bar.
The reality is more nuanced--the time to stop this legislation was by preventing it coming to vote in the senate. Typically the senate needs 60 votes to forcefully end debate[0], then merely majority to pass it. Once can disguise support for a bill by approving to end debate, then voting "Nay" given it will get the necessary 50 to be approved.
For JS 34 [1] Mitch McConnell (R, KY) limited debate to 10 minutes--I'm unclear from the transcript exactly how this was allowed. Richard Blumenthal (D, CT) offered resistance to limiting debate, and Kamala Harris (D, CA) and Patrick Leahy (D, VT) requested the role be called several times as a delaying tactic, but the limiting of debate went through.
Just prior to the vote, Brian Schatz (D, HI) offered some debate, but this is cosmetic given the known votes.
My read there is little to be gained by trying to legislate implementation power that has been ceded to the executive branch and the various agencies that are run by appointment, and therefore a costly filibuster and fight was not worth the time, effort and political mud.
The law provides a procedure for expedited consideration in the Senate. If the committee to which a joint resolution is referred has not reported it out within 20 calendar days after referral, it may be discharged from further consideration by a written petition of 30 Members of the Senate, at which point the measure is placed on the calendar, and it is in order at any time for a Senator to move to proceed to the joint resolution.[7] If the Senate agrees to the motion to proceed, debate on the floor is limited to 10 hours and no amendments to the resolution or motions to proceed to other business are in order, and so the Senate may pass the joint resolution with a simple majority.[7] A joint resolution of disapproval meeting certain criteria cannot be filibustered.[8]
Probably with good reason too, since if I'm understanding the US system correctly proposed regulations are created by unelected officials and can come into effect without any congressional vote whatsoever. Allowing filibustering of CRA votes would mean that regulations could be created despite the majority of the Senate and House strongly opposing them.
Not really -- rulemaking from the executive branch agencies is supposed to be what takes laws and implements them into concrete policy. The CRA is supposed to allow Congress the ability to void the rules that are perceived as going against the spirit of the law that was passed.
I haven't seen any reporting on a specific law that these rules were tied to, but I have seen references made to laws predating the public internet that mandate privacy on phone calls that.
People don't seem to like this comment, but that is a great link. Browsing around the site, I found this list of common English language errors: http://public.wsu.edu/~brians/errors/errors.html
Most of those are made up prescriptive guidelines, mixed in with some common misspellings. One grumpy person does not get to decide how the rest of us choose to use and evolve language. The "forceful" / "forcible" example is particularly inane and pedantic but hardly the worst on that page. I'll continue to say "being that", "ice tea", "center around", etc. as I please.
In linguistics the illustrative analogy is that prescriptivism is akin to an anthropologist entering into a foreign culture and rather than simply observing, they instruct the members of this culture on how to cook, dress, cut their hair, etc. Most modern dictionaries (including the Oxford English dictionary) take a descriptive approach to the study of language.
By that argument, we should all just be able to say and write whatever we want however we want to, even if it's technically or factually incorrect, like Humpty Dumpty or Donald Trump.
Why bother hewing to "elitist" rules of grammar and accepted spellings, being that it's just prescriptivism?
How does one decide objectively if something is just plain wrong, or merely prescriptive?
Case in point: "premises". So many people treat this singular noun as a plural and use horrors like "on-premise", which is so utterly wrong that it is painful for me to look at. What's worse is that "premise" is a real word and an entirely different thing and is most definitely not the singular of "premises".
This word came about (as many English words do) as a corruption of the Latin "praemissus", meaning something like "the aforementioned", and was used often in legal agreements for properties, and so became a word in itself that meant "the property".
Now we are corrupting it yet again, this time without the excuse of it being a different language, on the basis that "I'll say it however I please." People I have mentioned this to have told me that it is so difficult to get people to use the right word that they've just gone with "on-prem".
Now readers can take this comment as the rant of a "grammar nazi" or a pedant, but it wasn't meant that way, and I'll respond in advance with this: why is it not ok to identify something that is wrong? Because it's mere nitpicking?
Maybe so - but that's how matters devolve, over the decades, back to widespread ignorance and intolerance: one little oversight at a time.
Sorry, I didn't mean to get on the soapbox - sometimes it's just frustrating for those of us who are perhaps overly detail-oriented. But the world needs "pedantic" people like us more than it likes to admit.
The irony of course being that "being that" in the way you've used it is one of the examples of "incorrect English" given by the op. And you're using it to support the idea that these things are important...
> This word came about ... as a
> corruption ... without the excuse
> of it being a different language
You have some unorthodox ideas about how language came into being.
Point taken, I made a mistake, but I think you're so focused on picking apart the comment that you've ignored the meat of it. That's ok, I said my piece and expected it to be downvoted - and it was.
I'll upvote you (although technically HN doesn't like meta-commentary about upvotes / downvotes) because you made your point well.
I don't have an issue with every "error" on that page and I have no problem with style guidelines for writing, but the problem comes when the prescriptivists think that logic is on their side when they opine on subjective stylistic and dialectic issues. "Ice tea" is a perfect example. The author of that site argues that "iced tea is not literally made of ice, it simply is 'iced': has ice put into it.". Apparently he is unaware of English's enormous fondness for attributive nouns. By his reasoning, we should all be saying "appled pie", or "apple-infused pie", or something crazy like that.
I also have a more general problem with prescriptivism because frequently it is used to justify a certain type of racist and classist thought.
"Ice" is not a type of tea, but ice tea is a type of tea just as apple pie is a type of pie. I see no difference. Ice tea is tea with ice in it, apple pie is pie with apples in it.
Ok, even if I accept that, it's not a reason to prefer the term "iced tea" to "ice tea".
Ice algae are not made from ice, for example -- they are algae that are found in ice. The point is that in noun-noun compounding, the semantic relationship between the two nouns varies widely from case to case (far more widely than the distinction we are nitpicking over between "apple pie" and "ice(d) tea").
Which makes sense, because it's referring to a type of algae found in ice.
If there were a type of tea that only grew or were only found in ice, then it might make sense to call it 'ice tea'. However, that's not the case because it's regular tea that has been 'iced'.
The relationship between two nouns in a noun-noun compound is very flexible. Sometimes it means the head noun is made out of the attributive noun ("apple pie"), sometimes it means the head noun is found in the attributive noun ("ice algae"), sometimes it means something completely different (how about "ice axe"?). So, because that relationship is so flexible, it's just not absurd to consider that in the case of "ice tea" the relationship is that the head noun contains the attributive noun, as is exactly the case with "bubble tea" and many other NN compound examples.
And I will say again that the semantic relationship between "is made out of" and "contains" is so, so similar. Given the huge variety of acceptable semantic relationships between two nouns in a NN compound, it's really ridiculous to claim that "contains" is not acceptable whereas "is made out of" is, especially so when there are tons of examples of the "contains" relationship that staunch prescriptivists never object to (again, "bubble tea").
Sometimes it can be useful to expand the compound noun to see if it makes sense.
Apple pie:
- Pie found in apple(s)? Nope.
- Apple that is used to do something to pie! Hmm, no.
- Pie that is modified by an apple. No...
- Pie that is made with apple the primary ingredient? Yes
Ice algae:
- Algae found in ice? Bingo.
- Algae that is used to do something to ice? No.
- Algae that is modified by ice? Nope.
- Algae that is made with ice the primary ingredient? No again.
Ice axe:
- Axe found in ice? No.
- Axe that is used to do something to ice! Yes.
- Axe that is modified by ice? Definitely not.
- Axe that is made with ice the primary ingredient? No.
Finally...
Ice tea:
- Tea found in ice? No.
- Tea that is used to do something to ice? Not that I've heard of.
- Tea that is modified by ice? Mmm... it's not modified. It's still tea, only cold, not hot. Its temperature, a non-essential property of tea, has been modified. So wouldn't that be "iced tea", as in, "tea that is normally served hot but has been cooled down, namely, iced"?
- Tea that is made with ice the primary ingredient? No.
Bottom line: "ice tea" is ambiguous. "Iced tea" is not.
I regret I have honestly never heard of bubble tea (but I have heard of bubble gum), so I have no clue what it is, other than it has something to do with tea and bubble(s).
That is the relationship... "contains". There are many other examples in English of that relationship in NN compounds, and I guarantee you use them unconsciously without a second thought. You are also not getting my point, please reread my last post.
> Bottom line: "ice tea" is ambiguous. "Iced tea" is not.
This is how I know you and others in this thread have not spent a lot of time thinking about language. When has ambiguity ever prevented humans from using and understanding language? If you look at any piece of writing deeply, it is filled with an unimaginable amount of nuanced ambiguity. That's exactly why NLP is so hard.
But iced tea typically does not contain ice unless you add it. Buy a can of Lipton's Iced Tea - it's still iced tea, with no ice in it. But whatever, this is getting to angels on the head of a pin territory :)
I think you are making an unwarranted assumption about me. I have indeed spent a great deal of time thinking about language; I just have different thoughts, or points of view, about it. I have been very interested in etymology for a long time.
I neither claimed that natural language was capable of being entirely unambiguous, nor that people cannot communicate in the face of ambiguity. In fact, ambiguity in language allows for great artistic expression: humor, poetry, and other word play. So I agree with you on that point.
But holy crap, do we have to make it harder than necessary to communicate, when we aren't deliberately playing with words?
Surely you agree that much of the misery, pain, and suffering in this world of ours is due to avoidable language-related misunderstandings?
Sorry for being a bit rude. And yeah, this is the most hair-splitting argument I've ever been involved in on HN :)
My work in linguistics and NLP is strongly related to ambiguity, so I tend to see things in those terms and I do not see resolving ambiguity as an impediment to understanding language (for humans at least, but for computers it is an enormous problem). We'll agree to disagree!
What you call corruption, others call evolution. The fact of the matter is that there is no authority defining English. As a result, there is no way of objecctively deterimining if something is plain wrong. The best we can do is going by usage. You can try to influence usage, certainly, but some of those battles are simply not winnable.
Note that the prominent English dictionaries have usage panels that make judgements about whether the usage of a certain form of a word is sufficiently wide as their criteria for inclusion.
It's not that it is not ok to identify something as wrong, but you will need to accept that people will disagree with it, and that what is wrong to you now may very well have enough support in usage that the battle is already lost. When you then opt for comparisons to Trump, then it is not surprising that you get downvotes.
A lot of the "I'll say it however I please" is down to usage. I'll drag out my favourite example: "begs the question". It's my favourite because I didn't even know about the original meaning until I started seeing rants about how horrible the new meaning was. Do a search for it today, and the results are dominated by sites complaining about how awful the change is, and articles about it.
To date, I can recall only one instance where I've seen the original meaning used outside of such a rant. It's basically a lost battle, where people will often respond along the lines of "I'll say it however I please" for the simple reason that to most people the original meaning is entirely foreign because of its niche usage.
Usage panels, which often lags trends like this, for good reason, have in recent years started tipping towards the new usage for "begs the question", often marking the original form as "formal", because ultimately language is about communication, and you can not communicate effectively if you pretend the most common form doesn't exist. Here [1] is an article at Merriam Webster discussing the issue.
Yeah, look, I'm aware of language evolution (or devolution, as it may be), and that there's no formal authority for the language. I mean, there are already at least two major dialects of English (British and American), and there's enough of a difference between the two to cause problems for the unwary.
What I'm railing against is more that there appears to be little interest, in general, even to try to get things right. I see this not just in human language, but in business, software development, publishing, pretty much everywhere.
I'm really tired, so I'm not expressing myself as well as I should, so perhaps I should just wrap this up and get some sleep. An iPad is also not the best UI for writing on HN.
Thanks for your thoughtful and thought-provoking response.
> What I'm railing against is more that there appears to be little interest, in general, even to try to get things right.
Language is a means to an end. If my entire audience understands what I am saying, and is not put-off by how I say it, then I did, in fact, get things right.
Language is a set of tools. In the same way it can be annoying to watch someone hammer in a nail with the end of a wrench, it can be annoying to watch someone sort through their linguistic toolset, ignore the finely-honed implement meant for the job and grab another.
But when you're writing something down (especially for publication), you can not know your entire audience. Not using reasonable care in your written language communicates exactly that: that you care not for your audience.
Or that you care for a specific audience. I, for one, do not care about any audience members who get up in arms if I use "begs the question" in the sense of "raises the question", for example. The use has become so common, that I expect anyone who finds that offensive will not be worth the trouble for me to try to cater to.
You need to draw a line somewhere, or you will end up spending your life obsessing over unimportant details of what you write instead of actually communicating.
If your omission of what you assume is "unimportant detail" leads to your audience misunderstanding the "important" detail, you actually are failing to communicate.
It may also happen that the one person who you felt was not "worth the trouble" turns out to be someone who will be very important to you one day, like a potential business partner or investor, and who interprets your misuse of language as ignorance.
There's a difference between obsessing over unimportant detail and being thorough, IMHO.
Prescriptivism is dumb. I think everyone here gets that.
However, I like tools in my arsenal that enable me to express myself precisely. Prescriptivist rants often open my eyes to subtle shades of possible meaning that I otherwise would not have seen.
I don't quite agree with "typically". The numbers of filibusters (and cloture votes) has radically increased in recent history, but I'm not sure if you can really call it the default behavior yet.
Kensho | https://www.kensho.com/#/careers | Primarily: Boston, MA (Cambridge) and Washington DC. Case-by-case: New York (NYC) | ONSITE | FULL TIME
Kensho is applying machine learning and quantitative algorithms to timeseries, graph and unstructured data to make computer driven answers faster, more accessible, intuitive and beautiful.
-----
+ Software Engineers -- Front end, backend, infrastructure, APIs, apps, frameworks, performance, security, data wrangling. Machine learning skills a strong plus.
+ Machine Learning Engineers and Scientists -- You understand the math, the tools, and the implications of various algorithms and approaches. Software engineering skills a strong plus.
Who We Interview:
You stand out due to your work at a top technology company, research, and/or open source contributions.
Our Interview Process:
* We hope you'll share a project, paper or resume with us that highlights where you shine, with a short note so we can appreciate you as a person. Please say hi at jobs@kensho.com or https://www.kensho.com/#/careers
* As a small team, we'll reach back out to a few individuals to chat with a team member via phone, video or, if you are local, in person--to show and discuss your work, projects and code
* We may ask you to do a programming or data science challenge (<= 4 hours)
* We'll invite you to our Harvard Sq. headquarters to meet more of the team, where we hope you'll interview us too
* We'll discover we are peanut butter and jelly together, and wish we'd met sooner
* Having made you a non-exploding offer, we think you'll want to sign it
* You'll join us and have a lot of fun, get to play with fascinating data, algorithms and technology alongside delightful, hungry and creative people
* Something about being on a mission to change the world (hey, we're a start up)
Kensho | https://www.kensho.com/#/careers | Primarily: Boston, MA (Cambridge). Case-by-case: New York (NYC) | ONSITE | FULL TIME
Kensho is applying machine learning and quantitative algorithms to timeseries, graph and unstructured data to make computer driven insights faster, more accessible, intuitive and beautiful.
-----
+ Software Engineers -- Create beautiful web apps, dynamic visualizations, meaningful and non-flaky tests, composible and scalable infrastructure, cutting edge site reliability (SRE), neatly flexible operational frameworks, thoughtful APIs, practical yet robust security, and powerful frameworks for data analysis.
+ Machine Learning Engineers and Scientists -- Create advanced machine learning pipelines, NLP systems and new algorithmic techniques at scale using python, R or similar.
Who We Interview:
You stand out due to your work at a top technology company, research, and/or open source contributions.
Our Interview Process:
* We hope you'll share a project, paper or resume with us that highlights where you shine, with a short note so we can appreciate you as a person. Please say hi at jobs@kensho.com or https://www.kensho.com/#/careers
* As a small team, we'll reach back out to a few individuals to chat with a team member via phone, video or, if you are local, in person--to show and discuss your work, projects and code
* We may ask you to do a programming or data science challenge (<= 4 hours)
* We'll invite you to our Harvard Sq. headquarters to meet more of the team, where we hope you'll interview us too
* We'll discover we are peanut butter and jelly together, and wish we'd met sooner
* Having made you a non-exploding offer, we think you'll want to sign it
* You'll join us and have a lot of fun, get to play with fascinating data, algorithms and technology alongside delightful, hungry and creative people
* Something about being on a mission to change the world (hey, we're a start up)
Kensho | https://www.kensho.com/#/careers | Primarily: Boston, MA (Cambridge). Case-by-case: New York (NYC) | ONSITE | FULL TIME
Kensho is applying machine learning and quantitative modeling to timeseries, graph and unstructured data to make computer driven insights faster, more accessible, intuitive and beautiful.
-----
+ Software Engineers -- Create beautiful web apps, dynamic visualizations, meaningful and non-flaky tests, composible and scalable infrastructure, cutting edge site reliability (SRE), neatly flexible operational frameworks, thoughtful APIs, practical yet robust security, and powerful frameworks for data processing.
+ Machine Learning Engineers and Scientists -- Create advanced machine learning pipelines, NLP systems and new data modeling techniques at scale using python, R or similar.
Our Interview Process:
* We hope you'll share a project, paper or resume with us that highlights where you shine, with a short note so we can appreciate you as a person. Please say hi at jobs@kensho.com or https://www.kensho.com/#/careers
* As a small team, we'll reach back out to a few individuals to chat with a team member via phone, video or, if you are local, in person--to show and discuss your work, projects and code
* We may ask you to do a programming or data science challenge (<= 4 hours)
* We'll invite you to our Harvard Sq. headquarters to meet more of the team, where we hope you'll interview us too
* We'll discover we are peanut butter and jelly together, and wish we'd met sooner
* Having made you a non-exploding offer, we think you'll want to sign it
* You'll join us and have a lot of fun, get to play with fascinating data, models and technology alongside delightful, hungry and creative people
* Something about being on a mission to change the world (hey, we're a start up)
Who We Interview:
We scan your resume for at least one outlier experience, be it your undergraduate CS program, PhD program, open source contributions, research, publications, or previous/current employment.
Kensho is exploring new applications of machine learning on financial and unstructured data, making machine driven insights faster, more accessible, intuitive and beautiful. In all areas are are hiring hands-on jr, sr, lead and director levels of experience.
-----
* Machine Learning Engineers/Scientists -- Advanced machine learning, NLP and modeling techniques at scale with programming skills in python, R or similar. Think insightfully about data, models and problems and have an excellent nose for iteratively optimizing. Projects span timeseries and unstructured data.
* Back End Software Engineers -- Infrastructure, SRE, ops, security at scale. Please share a project repo with us. Thoughtful coding and high velocity are essential. Projects span high performance services, optimization, scaling and monitoring.
* Front End Application Engineers -- Front end/UI experts and enthusiasts who can lay out beautiful, intuitive, robust, and testable financial applications and workflow. Projects span simple to complex browser-based single-page-applications and other UIs.
-----
We will interview a few outlier candidates who share a project, repo, Jupyter notebook, portfolio or similar via jobs@kensho.com or https://www.kensho.com/#/careers
Kensho is exploring new applications of machine learning on financial and unstructured data, making machine driven insights faster, more accessible, intuitive and beautiful. We're small, hungry, and haven't made up our minds among sitting, standing or balance balls.
Machine Learning Engineers/Scientists:
Advanced machine learning, NLP or modeling techniques at scale. Notable research and data science experience is expected. Inside tip: Have multiple years of data science research, explain nuances of sophisticated models, think insightfully about data, and have an excellent nose for model optimization.
Software Engineers:
UI, infrastructure or SRE specialists. Inside tip: Share a project repo with us. High velocity problem-solving and thoughtful coding are essential.
Hiring process:
We're a small team who find traditional resumes unhelpful. We will interview a few demonstrably outlier candidates who share an impressive project, repo, Jupyter notebook, portfolio or similar via jobs@kensho.com or https://www.kensho.com/#/careers
We are making financial analysis faster, accessible, intuitive and beautiful through our partnerships with Goldman Sachs and CNBC. We're small, hungry, and have hoppity-hops in the office. To get our attention, share a project with us that shows:
Software Engineers:
UI, infrastructure or SRE specialists. Inside tip: High velocity problem-solving and coding are essential.
Machine Learning Engineers:
Advanced machine learning, NLP or modeling techniques at scale. Notable research and data science experience expected. Inside tip: Demonstrate multiple years of data science research, ability to explain nuances of sophisticated models and excellent ability to optimize.
UI Designers:
Your portfolio of data visualizations, workflows or UI designs. Inside tip: Make data beautiful, intuitive and informative.
Hiring process:
We're a small team who will interview very few candidates. We start with looking at the source for projects mentioned in your resume and/or cover letter. Then, depending on you, the role, and the projects you shared, we'll talk to you via phone/hang out/in person (if local). We'll likely do some live coding or design presentation, ideally on your computer, extending something you have created. We also have take-home challenges if you don't have deep enough projects to share, or maybe interviews aren't your thing. Lastly we bring you to Cambridge to interview us in person and go more in depth. Again, please bring along something you are passionate about and that you'd like to extend and discuss.
Kensho: Primarily Boston, MA (Cambridge). We also have offices in New York, NY (NYC), San Francisco, CA (SF), and Stamford, CT: FULL TIME, ONSITE
We are making financial analysis faster, accessible, intuitive and beautiful through our partnerships with Goldman Sachs and CNBC. We're small, hungry, and have hoppity-hops in the office. To get our attention, share a project with us that shows:
(Software Engineers) Innovation at any layer of the stack, but especially with javascript or infrastructure. Inside tip: High velocity problem-solving and coding are essential.
(Machine Learning Engineers) Advanced machine learning, NLP or modeling techniques at scale. Notable research and data science experience expected. Inside tip: Demonstrate multiple years of data science research, ability to explain nuances of sophisticated models and excellent ability to optimize.
(UI Designers) Your portfolio of data visualizations, workflows or UI designs. Inside tip: Make data beautiful, intuitive and informative.
Hiring process: We're a small team who will interview very few candidates. We all think interviewing is broken and have flexibility to adjust the process -- please show us where you shine! We start with looking at projects mentioned in your resume and/or cover letter. Then, depending on you, the role, and the projects you shared, we'll talk to you via phone/hang out/in person (if local). We'll likely do some live coding or design presentation, ideally on your computer, extending something you have created. We also have take-home challenges if you don't have deep enough projects to share, or maybe interviews aren't your thing. Lastly we bring you to Cambridge to interview us in person and go more in depth. Again, please bring along something you are passionate about and that you'd like to extend and discuss.
Kensho: Primarily Boston, MA (Cambridge). We also have offices in New York, NY (NYC), San Francisco, CA (SF), and Stamford, CT: FULL TIME ONSITE
We are making financial analysis faster, accessible, intuitive and beautiful through our partnerships with Goldman Sachs and CNBC. We're small, hungry, and have hoppy-hops in the office. To get our attention, share a project with us that shows:
(Software Engineers) Innovation at any layer of the stack, but especially with javascript or infrastructure. Inside tip: Your problem-solving and coding velocity are key
(Data Scientists) Advanced machine learning, NLP or modeling techniques at scale. Notable research and data science experience expected. Inside tip: Demonstrate multiple years of data science research, ability to explain nuances of sophisticated models and excellent ability to optimize.
(UI Designers) Your portfolio of data visualizations, workflows or UI designs. Inside tip: Make data beautiful, intuitive and informative.
First and most important: your internships and work experience, and what you accomplished during those jobs. They should tell a story of increasing and accelerating personal growth, learning, challenge and passion. If you can share personal or class projects, even better.
After your experiences, your degrees will be considered based on the number of years each typically requires, with early graduation and multiple majors being notable.
Put another way, if you don't have a PhD, the MS/MEng program is a tiebreaker compared to your experience and undergrad credentials.International students get a raw deal. The online masters will barely help you get a job or launch a career in the US. US universities appear to offer the chance to work for major US companies with a notable university (such as Georgia Tech) on your resume, only to feed their graduates into our broken immigration and work authorization system, H1-B indentured servitude and no replies from the countless companies that have an unspoken higher bar for those needing sponsorship.
To round out a few other contexts HN readers might experience:
If you are an international considering an on-campus MS/MEng, US universities are charging full price while giving you a credential of limited value and utility. Apply the same comments above but at a much higher price than GA Tech’s OMSCS.
If you are completing/just completed a less notable undergrad degree, paying for a masters program at an elite CS school (like GA Tech) is usually a bad deal. If it not a requirement for the positions you seek, it won't help your career chances much.
If you have an undergrad degree and your employer will pay/cover your MS/MEng at night/personal time (and that is your passion), awesome and go for it! It will be a lot of work and lost sleep to get everything out of the experience, but a lifelong investment in your growth and experience.
If you are completing/just completed a notable undergrad degree (tier-1, internationally recognized program), you don't need the masters. Feel free to get one for your learning, sense of self and building research connections while you ponder getting a PhD. The hiring and salary benefit will be very small--you are already the candidate every company wants to meet. If you decide to get a PhD, that will open some new doors but take 5+ years to get there.
At my previous company, we made it our forte and team passion to get authorization for employees--given a global pool of candidates and a hiring bar to match. I'm really proud of our effort here given the broken and unfair system. Sadly, many companies do not share this value or cannot justify the time, effort and expense, or cannot scale such a program to a larger number of employees across a less selective bar.