nlp tool{set}, pipeline, and custom datasets aimed at assisting job seeker / career wanters/needers/havers.
TL;DR -- Amazing sh*t happened. You should see it sometime, when it can be seen. We're working along mindfully. Stay safe -b
Update: A handful absolute breakthroughs have happened on the proprietary side of JobTalk source as of late
- we are currently reviewing our approach to community, freemium, and paid solutions.
- all solutions will provide fully-private, data-collection free, tools that empower users in their journey
- not just the bare minimum. paid options will provide additional products and services separate from JobTalk's suite specifically.
- JobTalk's name is DEFINITELY going to change lol.
- The core tech JobTalk now uses is named Freeform. Freeform is now fully decoupled from the aggregation of solutions engineered to compose objects specifically helpful to job talk and it's architecture...and those engineered and ever-evolving proprietary chunks of secret sparkle are able to be ingredients for other solutions born from the Freeform agent engine.
- The safety filters and latter mentioned Secret-Sparkle-chunks we've been developing are now designated under the name DebateClub.
- It is a truly exciting thought to demonstrate this tech to broader audiences as soon as January. Currently friends, fam, and folks I meet out and about that really fit the moment with this super-duper-prealpha testing momen. it's all in pieces, save the basic services for messaging.
- We are currently GPT-3 free and intend to stay non-dependent on sitch.
- I cannot tell you how insanely powerful this tech has been for me as a utility whilst building it – I never imagined things would evolve to this point of actual practical use to the point of signifance of being as radically life-changing as leaf blowers.
- But no matter how incredible,this new frontier of ability to augment one's day particularly demands strict observance of discipline and responsibility overall...as far as AI Safety is concerned. Establishing AI Safety guidelines with the approaching years in mind as well as the present is one simple thought toward this realistic glob of tech we really should be somewhat concerned with( still love it).
- We're consulting and working with friends in the ethics and safety spaces of machine learning...and what I honestly just like to now refer to as "the tech industry"... seeing as the industry appears to pose the greatest threat of catastrophic levels of carelessness in regards to ML but also overall.
Business Factory type stuff.
#TOOLS TODOs:
- installation script
- snippets:
- check for zshrc in $HOME
- zshrc path update for tools dir
#probs a zsh path var.. code below is likely overkill or needs to test the path given test ls $ZSHCONFIGPATH/.zshrc && ( cp $ZSHCONFIGPATH/.zshrc $ZSHCONFIGPATH/zshrc.ttbak[date] $ZSHCONFIGPATH/.zshrc echo '[export path]' >> $ZSHCONFIGPATH/.zshrc ) || ".zshrc not found in path $ZSHCONFIGPATH.
Please add following code to your .zshrc:
...."
- dataprep
- employ nltk or other nlp framework options for data processing
- update regex's to leverage hyperscan in * manner relevant to* employed technique for processing text. e.g. *sh scripts vs py vs binaries