935 karma · joined March 7, 2012
For applying to jobs, consider being creative in your application. Can you speak directly to the hiring manager? Can you make your application stand out somehow?
The best engineers I've hired have usually been ones with the drive and without formal education. Best of luck.
This is the main reason why big companies are never allowed to fail. Why banks and all the traditional giant establishments continue to live. It's also why publicly listed companies that have been dividend paying for some time would rather layoff employees when growth prospects are scarce than to rather maintain them on payroll.
So if your company's stock is ever owned by government/pension funds, you'll have a pretty good argument for getting bailed out until you find your next trick.
Spoke to an experienced engineer who used to lead NLP at MSFT and same comment. NLP models are already fantastic and it isn’t very hard to build a smart chatbot. The implementations these days are just very poor because they are not well thought out from a user perspective.
Would love to chat if you have further thoughts around the subject - there's a ton of problems we're looking to tackle in the space and would be good to get input.
There’s so much hype right now that silly money is just being thrown at new tech without fully understanding the value you’re getting.
2. Handcoding to begin with is fine but as you start to scale the number of production models and actually productionalize the model at scale, it’s unfeasible and leads to plenty of maintenance issues. There are a few model infrastructure tools that help with this but again, many are homegrown because the market is still new. Algorithmia, Seldon are pretty good starts.
3. Rarely use serving options provided as the challenge is integrating it with the rest of engineering. Service monitoring gets handled by different teams.
4. Depends on the industry and usecase. Again integrating and maintenance comes into play. Go/Cortex might make sense but a lot of companies leverage Spark so Scala/Java could be the choice for production models.
5. We’re working on creating this recipe for enterprises. I believe Seldon (open source) might contain this capability. The challenge as you pointed out is ensuring things don’t break!
Where we see and place all our analysis efforts are on companies like Google, Amazon, Apple, etc. From what we can see with Alibaba, since they own the contact points for the both the customer and retailer, they have data to gauge financing needs and more importantly risk.
You see banks like RBC starting the conversation earlier in the journey with their investments in various startup spinoffs where it's focused on the 'why' behind the financing (i.e. finding a home, buying a car, starting a business, etc.) before becoming involved in the financing portion later down the line.
First mover advantages are less becoming of an advantage because of how fast information travels. The other obstacle with being first movers is the amount of effort needed to first educate the market on the new way of doing things. Traditionally, first movers had an advantage because followers couldn’t move as fast and wasn’t able to piggy back on learnings as easy as they can now.
You’d be able to track long term and short term data distribution changes so you have an idea of when your model could be underperforming and will require retraining.
Would love to learn more about your project.
We are indeed hoping to start off with a self-service offering mainly geared towards AI startups or data scientists at smaller companies.