TFIP IBF Cloud Computing

[Kingdom]

Any updates for this week? After my interviews in June, I haven receive any notifications.

Similarly, not for me yet after my last interview with Citi.

 

[Dalias]

Just finished my interview with the prospective employer, no technical questions asked, more of a informal chat to understand why are you making the switch, and the driving factor behind and the associated risks of doing so for both them and us.

They also highlighted the difficulty of the individual coming in to a new area, where they already have a established team of experts and it may be difficult to "teach a old dog, new tricks" where they already have a ready pool of younger candidates which are skilled available.

While they are generally supportive of the scheme, but it seems that from the practical aspect of it, they have some hesitation with the fear of failure and therefore wasting everyone's time which is notably understandable. From my own feel and feedback from another candidate, seems like they are doing this solely to show support for the initiative and would still prefer someone who is a plug and play model.

 

[j4yp33]

Just finished my interview with the prospective employer, no technical questions asked, more of a informal chat to understand why are you making the switch, and the driving factor behind and the associated risks of doing so for both them and us.

They also highlighted the difficulty of the individual coming in to a new area, where they already have a established team of experts and it may be difficult to "teach a old dog, new tricks" where they already have a ready pool of younger candidates which are skilled available.

While they are generally supportive of the scheme, but it seems that from the practical aspect of it, they have some hesitation with the fear of failure and therefore wasting everyone's time which is notably understandable. From my own feel and feedback from another candidate, seems like they are doing this solely to show support for the initiative and would still prefer someone who is a plug and play model.

Aia? Which other tracks / fi did you apply to?

 

[Dalias]

Aia? Which other tracks / fi did you apply to?

you are right, and one more FI (i might have forgotten, either UOB or Citi). but no calls from FI.

Did you get interviewed with aia ?

 

[j4yp33]

you are right, and one more FI (i might have forgotten, either UOB or Citi). but no calls from FI.

Did you get interviewed with aia ?

Applied for citi, Stan chart, aia for cc. Only got call from citi

 

[Bombermankid]

Hope next week onwards we can get our result.

Get result also no use if lee-jected ? Unless they tell you they won't send your profiles to the remaining FIs anymore. If not, then the wait just restarts again.

 

[Bombermankid]

If rejected early, at least no need waste time waiting?

Let's say you applied UOB and Citi, then end of this week you know you are rejected by both.

Then Ibf tells you that they will send your resume to the remaining FIs.. So you won't want to go to the other FIs even if there is still a chance?

I believe most will take up this program no matter which FI it is.. So if you got rejected by your first two choice this week.. Then wouldn't you still have to wait again to hear from the remaining FIs.

Unless of course, if you are not interested in the remaining FIs at all.. Or if Ibf is not going to send the rejected profiles to the remaining FIs due to huge volume of applications.

 

[Kingdom]

IMHO, what i'm weighing in my mind is the opportunity cost I have to pay while waiting for IBF's reply. I noticed there are a lot more good offers in the market now compared to the past months due to SGUnited initiative. By the time, the so called "standby" batch be informed "thank you for your recent interest in... we regret to inform you that..." it may probably be somewhere in August I guess. Then, we will be loosing this whole month to waiting for something that didn't get to materialize... Personally, I have a current job, so still not that bad to wait for their reply. I also believe those of us that are unfortunately not being further selected will be first informed amongst all. Let's hope for the best yet in parallel be wise enough to prepare for the worst outcome. If one is prepared, one may not be caught in dismay when things do not unfold the way we would like it to be. Just my 2 cents worth.

 

[Black_Gate]

During interview i asked. The manager told me at least need to wait till mid or end of Aug to know all the result.

And I think probably will delay to Sep at least

 

[j4yp33]

IMHO, what i'm weighing in my mind is the opportunity cost I have to pay while waiting for IBF's reply. I noticed there are a lot more good offers in the market now compared to the past months due to SGUnited initiative. By the time, the so called "standby" batch be informed "thank you for your recent interest in... we regret to inform you that..." it may probably be somewhere in August I guess. Then, we will be loosing this whole month to waiting for something that didn't get to materialize... Personally, I have a current job, so still not that bad to wait for their reply. I also believe those of us that are unfortunately not being further selected will be first informed amongst all. Let's hope for the best yet in parallel be wise enough to prepare for the worst outcome. If one is prepared, one may not be caught in dismay when things do not unfold the way we would like it to be. Just my 2 cents worth.

Agree. This tfip was my backup plan, but looks like will be my main plan now. Jobless since March, but still can sustain thankfully.

 

[ktsim]

Me out of job since mid April, hopefully most people here still have a job though.

I'm also out of job since beginning of Apr. No success in landing on a job yet so giving this TFIP a try.

 

[Dalias]

Anyone here has been called up by AIA and already received an offer from them?

Just checking because i went through the interview and didn't hear back until now. I was only shortlisted for this path, and the interviewer told me that likely only 1 or max 2 slots for him to take in.

 

[havetheveryfun]

some FI i heard have a quota of 4 people. Did not expect that AIA only have a maximum of 2 slots. That is really a tight competition there.

not surprised at all, so you did well in getting an offer

 

[havetheveryfun]

Well thanks for the praise. But I do have a bigger concern over my ability to be converted. Cause at that point the playing field will be much more intense as I will be competing with much more experienced colleague for the perm position.

well.. i would say some of the FIs take part in this program without having any intention to convert anybody into perm staff. After all they can just continue to take part in this program every year and have many candidates to choose from again every year, as long as IBF and the government continue with this.

But the 2 years experience and the courses you take and some of the certs you get along with it will definitely make it easier to find the next job in that particular industry after the 2 years, after all this is what the core of the program is still ultimately about.

 

[datageek]

side track abit from the tfip interview-offer chatter.

so far, i have been working on statistics, linear algebra, ml stuff and analytics using r and python and the array of libraries that come with them. pretty data science learning in a pretty sandboxed environment..

took a short break from work, went googling on container docker deployment and stumbled onto a 'whole new world' out there called "architecture of a machine learning pipeline system".

https://towardsdatascience.com/architecting-a-machine-learning-pipeline-a847f094d1c7

after taking a 5 mins glean on this write-up, i have come to a fuller picture of the entire software architecture on the ml pipeline system. it cuts across so many areas in software engineering, data storage, processor, server ... like the whole eco system from start to middle to end.

forgive me for my ignorance, i am not working in the tech industry. all these are pretty new to me. haha.

anyone knows this whole spectrum of tech well? if you do can you kindly elaborate and summaries for me and some bros out there who are interested?

thank you

 

[Black_Gate]

side track abit from the tfip interview-offer chatter.

so far, i have been working on statistics, linear algebra, ml stuff and analytics using r and python and the array of libraries that come with them. pretty data science learning in a pretty sandboxed environment..

took a short break from work, went googling on container docker deployment and stumbled onto a 'whole new world' out there called "architecture of a machine learning pipeline system".

https://towardsdatascience.com/architecting-a-machine-learning-pipeline-a847f094d1c7

after taking a 5 mins glean on this write-up, i have come to a fuller picture of the entire software architecture on the ml pipeline system. it cuts across so many areas in software engineering, data storage, processor, server ... like the whole eco system from start to middle to end.

forgive me for my ignorance, i am not working in the tech industry. all these are pretty new to me. haha.

anyone knows this whole spectrum of tech well? if you do can you kindly elaborate and summaries for me and some bros out there who are interested?

thank you

To use AI in real world application actually involve 2 big part training and inference.

AI engineer mostly doing the training part AKA model. Basically you are adujsuting model parameter to make sure you get what you want.
And whole training process repeat until you get high accuracy output.

Then you go to inference part when you put the model into real application like real car (for autonomous vehicle) and let him run on real road.
If it fail to make correct decision in real road we go back to training part.

training -> inference -> training ......


Now the most famous ML framework are from google call TensorFlow. They make it open source. so it is very easy. You just call the function.

Few 10 lines of code you ard can do the training. Then export the model to your phone/PC app and you can start the inference part. Let AI identify the real world picture.

If you are just simple AI engineer, you will spend a lot of time to do parameter adjustment and highlight obvisou feature for your baby (like photoshop) . If you are Phd level AI engineer, then that is different level, you probably will spend a lot of time to create new model methodology and so on and it is very very complicated.

 

[datageek]

To use AI in real world application actually involve 2 big part training and inference.

AI engineer mostly doing the training part AKA model. Basically you are adujsuting model parameter to make sure you get what you want.
And whole training process repeat until you get high accuracy output.

Then you go to inference part when you put the model into real application like real car (for autonomous vehicle) and let him run on real road.
If it fail to make correct decision in real road we go back to training part.

training -> inference -> training ......


Now the most famous ML framework are from google call TensorFlow. They make it open source. so it is very easy. You just call the function.

Few 10 lines of code you ard can do the training. Then export the model to your phone/PC app and you can start the inference part. Let AI identify the real world picture.

If you are just simple AI engineer, you will spend a lot of time to do parameter adjustment and highlight obvisou feature for your baby (like photoshop) . If you are Phd level AI engineer, then that is different level, you probably will spend a lot of time to create new model methodology and so on and it is very very complicated.

hey blackgate, thanks for taking the time to post a long and detail one. thank you for putting it across in a simple and easy to understand manner.

the analogy you shared reminds me of the race car model that AWS use to teach reinforcement learning.

it seems that the spectrum of AI is very wide, from data engineering to ml engineering to data science stack. as such, looking at the analytics and AI tracks of TFIP, the data analytics track is biased towards the data science stack and the AI tracks towards the data engineering and ml engineering stacks. these 3 stacks will take a long time to master and at the same time, the 3 stacks are constantly evolving and with new frameworks. everything in a flux all the time.

seems like you are working in a tech role now. care to share more about your tech stack and experience? (looking forward to learn not just from course but tech guys like you)

 

[datageek]

Just curious, for those taking AI track, do they need to perform both the task for DA and AI? Cause i find that it is difficult to focus on multiple specializations.

For my case, i am focused on cloud computing and I only implement the best security practice for the infrastructure, instead of dealing with multiple specialty such as cyber-security.

hi jgyy,

ai is a pretty generic terminology. as far as my understanding and research, it covers a wide spectrum, from data engineering to ml engineering and some of data science stuff.

the ai track emphasis should be on data engineering and ml engineering in which you create and build data pipeline to train and build your models (including hyper-parameter tuning). deploy your models in docker containers. use of cloud computing infrastructure such as azure or in-house cloud. use data processing technologies such as nosql, apache hadoop and/or spark.

as such, a strong background in programming and technologies (comp science) usually fits the job scope.

hope it helps.

 

[Dalias]

Just curious, for those taking AI track, do they need to perform both the task for DA and AI? Cause i find that it is difficult to focus on multiple specializations.

For my case, i am focused on cloud computing and I only implement the best security practice for the infrastructure, instead of dealing with multiple specialty such as cyber-security.

@jgyy1990 - just curious.. you sound like you worked in the Tech industry and in cloud technology.. is that the right understanding?

 

[Dalias]

I did not graduate with CS degree, my first 4 years after graduation is working in shipyard. Only recently my last job before contract ends got touch on cloud.

ok cool.. sounds like a techie to me. haha.. btw.. i just saw the offer come in for the role in CC for aia..