Brett Rants

Job Offer

Well, that's a bit optimistic. It was a job lead that fell through. On the plus side, my time spent considering the lead has convinced me that I am interested in obtaining an actual offer.

random geometric lines highlighting the architectural elements in a ceiling, it has a series of horizontal lines, and a few circle elements from an alarm bell and smoke detectors or the like

The Job I Want

Let's assume you've landed on this page because you are considering extending me an offer for employment.

This is what I want.

The Exchange

Like I can remember the exact exchange at this point. For some reason, I got two leads that week... and the one said part-time remote, so I said 'Why not?'

Of course, that's not clear.

I've got an ageing resume (or two) somewhere in the ether (call it the Internet). I really should update them. But then, folks sending out inquiries should really read my listings.

See, I've got 25 years experience doing... stuff. Does it really matter what? Maybe it does, as it was not working with computers. I've only recently become interested in computers. I would guestimate that I have 6-8 years of experience working with code at this point. But as the starting date was far from momentous, I'm not certain. Splitting the difference and calling it seven years is my best guess, as the six year estimate comes from how old this website is. But by then, I'd already come to believe:
Real Programmers have their own websites.
So, I'd already started to think of myself as a programmer.

Anyway, at my high-point, I had three websites. But there was no point in paying for the additional domains, so now I am down to the one... this one.

That doesn't matter. What's pertinent is that my 25 years of experience and the fact that I know Python get conflated and mixed together, so I get inquiries that overshoot my programming abilities on a regular basis. Perhaps, I should make that clearer in my profiles. But then, it already is pretty darn clear. So if it feels like I'm messing with you (Mr Recruiter, Man), maybe it's because I know you haven't bothered to spend thirty seconds reading my profile.

Anyhow:
  1. They inquired as to interest.
  2. I asked who they were.
  3. They told me.
  4. I researched the company.
  5. I sent something similar to the following.
    • The version I sent them was not obfuscated.
  6. And I never heard back.
    • So, sad.

My Formal Response

Dear Mr Recruiter,

The Property looks interesting.

I'm in the middle of reading the Sales Brochures I found online.

Per This Sales Brochure, I will assume Python npl (or similar library) is used under the hood to parse text and title features. And if I were using The Property, rather than a simple note that The Wording was significant, I would want to know exactly why. Perhaps this is already implemented. I did not note anything in this regard in this particular white paper, though. Anyhow, digging deeper into this issue (or working for the team that does) could be fun. I don't know the npl library. But I would like to. So, fun. Um, I also should note, that by 'I don't know the npl library' I mean I've never used it for anything past a toy project. I have read the complete online documentation, however. And I found it quite impressive (i.e. clear and accessible).

Per A Different Sales Brochure (I forget the exact name), malicious code detection seemed to be a central item of concern and the Sales Brochure indicated The Property was moving beyond (or hoped to move beyond) Hash Fingerprinting (or whatever exactly verbatim code identification is called). Binary Analysis (malicious code pattern analysis) could be fun (see above). And much to my chagrin, I have been unable to find much of meaning on the Internet regarding this topic. I can see how this would be an ideal use-case for your Solution.

Finally, if The Wording is predictive of Economic Success, I would have to assume the content of an image would do the same. I am no expert, but I am really comfortable with images.
{And the more I think about it (after the fact, way after the fact), the more interesting this particular topic seems.}
As to my skills, I am proficient (or so I would have you believe) in Python (2 or 3, but I use 2 almost exclusively for my own stuff). I use Numpy routinely. I have used Pandas, but the library has not been a good fit for my use cases. As to This or That {let's not be too specific}, I am sure I've read papers on both but but use neither. For instance, I would guess This is used for the front end (I would have guessed More Name Dropping out of the blue, not being familiar with that either) and is a paired down Front End Solution. I think it used to be called Something Else {so, I know stuff, see}. Either way, a quick look at My Projects will show that I just don't use it or anything close to it. That Thing {of the aforementioned This and That fame} (I believe) is A Technical Explanation Here {that I think I nailed}. For my use cases, A Much Simpler Solution is the Superior Alternative, so I have no working knowledge. More Tooling (once again, I assume) is a More Complete Explanation that I have never used. I do not know anything about Most of the Stuff Listed in the Job Description.

Also, as stated in my original email, I know (next to) nothing about statistics. For my own use cases, I would use Python over R. And I have no intuition as to what Standard Deviation really means {though, I do have a passable non-intuitive understanding -- the mathematicians in the audience will (hopefully) know what I mean}. Though, I think I have fairly good grasp on correlation. Please see the links below if interesting in trying to parse my mathematical knowledge.


Resumes can be found at Resumes.

Code samples (I like the Tic Tac Toe one the best) can be found at Code Samples.

Also, more at the base directory: More Code Samples.

And really, the result of code (if not the code itself) is all over the place, for instance: Graphing Examples

This one includes the implementation of a fairly difficult algorithm (so, mathematical acumen): Lottery Analysis

Oh, hey. And the job description talks about physics, so: Further Proof That I Am An Idiot

To reiterate the sort of thing I am looking for: Python, Remote, +/- 10hrs/week.

My phone number is A BIG NOPE {but it doesn't really matter as...}
My ringer will be off (it is always off), but text me (preferably) or leave a message and I will call back. I will not be checking this email account for the next two weeks, so do not expect a response on this front.

I hope to hear from you.

Sincerely, My Name Redacted For Funsies {because if you don't already know who I am or can't figure it out else-wise, you are a hopeless case}.

Like I said, I never heard back from them. I'm sure it was because they were looking for someone with more experience, I provided too much information and they didn't want to pursue someone with so many technical holes, and/or my attitude shined through.

Who knows?

I am not upset about the lack of a job offer. I would have loved an offer. But that is neither here nor there. What's important at this juncture is for you all to understand that I am not grinding some ax, trying to kick This Company (hence, the redactions), or get revenge in some roundabout manner. I am simply having fun, doing my thing. And a quick purview of my site will show that rants like this are exactly my type of thing.

Also, I want to sort of work through what went wrong... not that I have much to go on.

And finally, I've got a ton of notes that I made during the process (my reading of all those Sale Brochures, as I did my due diligence). And this write-up salvages those notes and kills the chance of getting hired by anyone else (like, ever), all at the same time.


My Notes

If we (or perhaps, I is more accurate, as I am sure they called someone) had ever gotten to the telephone interview stage, this (as in, the content of these notes) is what I would have brought to the conversation. And bringing this level of input (high or low, good or bad, I will let you decide) is what I bring to any project, paid or otherwise... you know, in addition to mind-boggling good looks... and humbleness. I am so much more humble than the average bear, it actually startles me sometimes. I just look at myself in the mirror and think 'Gads! What a looker!' I mean, the world is such a better place with me in it, especially after taking into account that I don't rub my superiority into other peoples faces.

Also, I'm fairly well convinced that I am an idiot. So, there you are.

Anyhow, my notes...


Online Resumes


I need to updates these. But you know, that will never be a high priority.


Remote - Python - Part Time


But we have been over this.


Their Website

As in, this would have been my feedback regarding their website if they had ever asked... and I felt like telling the truth instead of being polite, as I can be polite. After all, it's amazingly easy not to say anything. But I hope you will (at the very least) grant me that a rant (blog or whatever you want to call this) wherein the author says nothing is entirely pointless.

Anyhow, in regards to their website:

Cookies
I received endless cookie requests. Um, at some point their server should have realized that I wasn't interested and that I had disabled cookies on my computer.

No Anonymous Requests
You would not believe (or maybe you would) how much marketing material was hidden behind an email wall (in that they wanted my email address prior to allowing me access). Um, no. I come from a tradition (as I believe most programmers do at this point in time) of Freedom! Give me Libre de Software... or give me the closest Open Source Alternative! Actually, I am not a die-hard. And the code I would have been working on was proprietary. But I can't see (me, myself, and I) bothering to dig very deeply into their marketing material sans job interest. It was not the easiest site to navigate anonymously.

PC PR
Most of the documentation I uncovered was targeted at Middle Management... or at least, it was Marketing Oriented. I am a Tech Guy. I want to get under the hood as fast as possible. I could care less about Best Practices... or any other Business School Buzz Words. Seriously, I gagged on the Buzz Words.

Why Not Open Source?
Um, there were Open Source alternatives to This Solution. Knowing this, I wanted to know PDQ (as in Pretty Darn Quickly) why those Free Solutions were not The Superior Solution. And they had their answer, but it would have been very difficult for me to quantify (the benefits of their Solution) without getting a quote from them... which means, Their Solution was (way-way-way) out of my price range. And therefore, I am not the target customer.

So, no harm. No foul.

And therefore, my opinion regarding their website ain't worth a can of beans.


Machine Learning

If you have gotten this far, I do not mind telling you that The Solution was in the realm of AI, Machine Learning, Deep Learning Neural Networks, and all that jazz. And thus (mainly so I don't feel the need to write another page about this subject and can get it out of my system, right here and now), let's talk about (what I presume is the cutting edge of) Applied Statistics for a moment (as it relates to The Solution, but of course).

Tuning the System
I don't know if this was the exact wording, but it has interesting implications: it means the power of Subjective Human Insight is not to be overlooked. Or in other words, the problem is not yet solved. We are just at the beginning, as the Machines cannot tune themselves.

The Problem Set
It was actually more difficult than I would have imagined (perhaps through lack of imagination, an over imagination, or lack of familiarity with the problem set), but I would have thought the Solution Domain (so, the Range of Problems, unless I have that backwards) would have been easier for me to recognize. I mean, at the end of all that Marketing Material, I got it. But towards the beginning, I was imagining bringing my own Data Sets to any Job Interview to Stress Test the System.

So, my Problem Sets... as in, these are the Problem Sets currently on my computer, waiting patiently for me to bother to look at them:
The Solution (as near as I can tell) would have choked on (or really, just ignored) all of these inputs, which begs the question: What is the nature (i.e. format) of the Data Set the The Solution can ingest?

And as I just said (so, listen to the beat), I do not think their desired input format matches (the format of) any of my Data Sets... which goes a long way towards explaining why I have yet to dip my toes into the AI Well.

The Solution
On the other hand, they were not trying to solve my problem, but their own... and they likely had a best of class solution.

With Friends Like These
They had a fair number of 'Partners'. Though, to be honest, most of the 'Partners' (and I will keep on using quotes) looked like competitors (note the lack of quotes). I wonder how many of these 'Partners' were... um, investment related... funded by the same folks, and/or tied more by outside money interests than inherent need. I don't care to look into it. But I think it is a serious question.

More Friends
In addition to 'Partners' there were (what I will call) flat out dependencies. And it is a wonderful bit of marketing to call dependencies assets.
'See, they solve a problem.'
'But I thought you were solving the problem?'
'Well, in truth, our solution has problems of it's own.'
The above exchange being both fictional and unfair. But to any joke (no matter how inside it may be) there is always some grain of truth.


Obscuration

And I think that is all the specifics I wish to go into, though I may revisit that whole AI thing, again (in some other post). In the meantime, I give you one of those lists that is almost completely devoid of meaning, but for whatever reason seems to fulfil a need for me.
{While reading the first two Marketing Pamphlets, I was wondering about my interest in any job. I was like, 'You don't respect them. Why would you want to work for them?' But then, I got to the meat and potatoes of The Solution and it was truly Best in Class. So, I was willing to blame the idiotic (in my opinion) marketing on the VCs. And assume (for the time being), the code itself was solid... that working there could be fun.}


Possible Projects

So, like these are some other (AI, Machine Learning focused) projects that I might wish to pursue at some future date... likely after I run out of things to say (Ha!) and need an involved coding project to pass the time... or someone wishes to give me a job pursuing one of these. So, one could look at the following as some sort of Job Project Wish List.

Text Parsing
I believe I mentioned nlp (Natural Language Parser, a Python Library) previously. Someday, I'll find a project for it... and in this context, that would mean using nlp as a pre-parser for a ML Toolchain... forgive me if I cannot be bothered to come up with a few more buzz words to tag onto the end of that sentence. Of course, if I did (or could), then I would have my project.

Binary Thumbprint
I like images. And a few years ago, I found a write-up (a published research project) in which different code segments were visualizing by encoding the zeroes (in the base binary) as black and the ones as white and transforming the lot into black and white images. It turns out, certain types of programs have different looks... but only to a limited extent. Still, these sort of subtle differences are exactly where Neural Networks shine.

GPS Movement
I track my GPS Movement... and so does the phone company. But that's not really the point. Humans move. And humans move in different ways at different times based on their intent (what they are doing... and personality differences, but of course). So, like, one could track customers via in-store cameras and likely be able to split said customers into buyers and non-buyers (big spenders, one off purchasers, browsers, and much more importantly shoplifters) by the way they move. Throw in Body Movements and you've got yourself a wrap. Outside of detecting nefarious intent, the real benefit is in giving individual shoppers the type of assistance (and/or shopping experience) that they want. Sometimes, I want to be talked to; and sometimes, I do not.


The Great Idea

Mathematics will always be a Hobby Subject for me. It's not a main occupation. And it will never be how I shine. I'm not willing to put in the work. So, for all I know this may be standard practise. Anyhow, after looking at (or more accurately, thinking about) some Classification Problems (e.g. is this customer a shopper, thief, or whatever), I concluded that a second tier test would be appropriate.

Um, how to describe this.

No model is 100% accurate.

And no model will ever be 100% accurate.

But in certain models, there will be factors of over-powering importance.

Here's a good example (the facts of which may be wrong). Heart Attacks are predicted by being Over Weight, Smoking, Drinking, and so on, and so forth... until one gets down to the Salt in One's diet (of slight importance, considering the rest). But the greatest predictor of Heart Attacks (I would guess) is Age. So, given a person is this Old and Weighs this Much, the trick (the trick I am proposing) would be to make a new model with those two factors as a given, where all the test subjects are this Age and this weight.

So, like, Is this Airline Passenger a Terrorist Risk? becomes a different question based on the Airport of Origin.

Or in other words, Given a High Probability Hit, the Test Population is reduced to only those who match the Major Criteria (whatever that means) and a New Model is Constructed against which the Subject is Tested.

Given a tentative hit, refine the criteria.

And then, let me reiterate, no model (for any complex subject) is 100% accurate. And this in itself implies missing features.

Anyhow, this is part of the value-added I think I bring. Of course, at ten hours a week (and a non-mathematician at that), it's likely not the value-added I will ever be hired to bring.

Oh, well.

I shall write rants, instead.

And gift my insights to the world at large.

And/or those robots smart enough to crawl my site.


The Debriefing

I haven't heard back from them.
Waiting...

Waiting...
I guess I am not going to.

Phone Interview
I did not get a phone interview. I would have liked to have one. I desire to have some in the future... you know, at least one, as I think that's a prerequisite to getting a job. But then, I also don't need to talk to someone on the phone just to determine that we are not even close to being a match.
What? Ten hours a week? Are you insane? We were going to ask for forty... and demand eighty.
I don't need to waste either of our times.

But given a phone interview, I think The Company should at some point ask for feedback. As in, I would imagine requesting feedback from interviewees would be 'Best Practices'. And if it's good enough for them, it's good enough for me. I mean, granted, a rant like this is not for everyone. And I don't really want a bunch of negative criticism. Besides, those who would criticize my work (as a lot, as a whole) are not really the sort I want to work for. Thus, I can see A Company being selective in whose feedback they solicited (and in fact, only the feedback of possible customers really matters). Still, The Website I looked at was, um, not overly friendly or inviting.

Of course, if they don't care about my point of view, maybe we don't have a match.

See how that works?

Play The Game
Now, I won't do this, as this is not who I am. But I have thought about trying to keep score somehow: maximize leads (1pt per email), phone calls (10pts per), interviews (100pts per), and so forth... leading to job offers, but of course (at 1000pts per). I mean, I won't do anything like this, because my time is worth something and I have no need to waste my effort collecting Fake Interview Points. But I would like to get further along in the Hiring Process. And incentivizing my progress might help in that regard... you know, by making a game of it... or writing up a rant or two... such as this.

New Resume
Well, I don't think I need a new resume. What I need is a page (not this one) that explains what I want and what I have to offer: Python, 10hours/week, Remote, and more attitude than you can shake a stick at. Clearly, I am going to have to work on my presentation.


Alpha Channel

Finally, I am not convinced that I want a job or could work one even if I was hired. So, I thought I would start a project and try to work it like a job, try to be disciplined, and work it ten hours a week. The name of that project is _Alpha_Channel_ and will be a series of webpages devoted to command line image manipulation.

I've spent enough time, here.

It's time to start working on that.


next Brett Rants entry

Home Brett Rants Index

It is going to be difficult making _Alpha_Channel_ a priority.

© copyright 2019 Brett Paufler
paufler.net@gmail.com