 during your course of data science, one thing that you will have to do again and again, and you have to do it in different areas or different forms and shapes, you have to come across it every day, every time there are two things in it, one is ETL and one is ELT, the purpose of both is the same but depending upon which system you use, what skills you have, which system you are working on, what is the nature of the data, then you use either one of them or sometimes both of them, again it will depend upon your business requirements and the situation and the organization which you are handling basically you see that there is a little difference between the two, ETL means extract, T stands for transformer, L stands for load, ETL means extract, transform and load in this we do that when we took data from source and we took data wrangling, data quality, data completeness where we discussed, or you have noted in two terms, one is enrichment and one is transformation basically when you take data from source, then it is enriched or transformed or complete, then you save it in your target database so from the source system you took data, then there is an enrichment or transformation, like we said, I think now we should not talk more about the example, we shared that one sheet with you, then in the other two multiple slides I discussed in the previous module, I have given you a couple of examples and I believe that now your basic understanding is about data extraction, then we have to enrich the transformation on it and then we have to store the database in it, I am sure that you are at par with these terminology so these are the differences, the first one is ETL, the second is ELT, in ELT you extract and load it in the target database and then you transform it, it will depend on the situation, now if you are working on real time data which is near real time so there you have to transform some time situation or you have so little time that you cannot transform it, you cannot enrich it and this is not your business requirement, you say that I will take it as it is in the data, after that I will take it in the target database then whatever process I have to do, I will talk about it after the transformation, so you can manage these two situations in this manner now the transformation is basically, if I give you an example of the banking industry, then if you are getting transaction from ATM, online banking, you know, branch less banking, now there are so many things that are going on within the banking industry, you note that now everything is on your mobile phone this means that your telecom industry and banking industry is getting merged as far as the technology is concerned on the other hand, if you look at your telecom operator, they have also made their own bank, you have this mobile cash, mobile wallet that is the form of banking so this means that both the industries and their technologies are getting merged so the bank is giving you the facility of mobile banking, the mobile operator is giving you the transaction of cash or even the bank is providing you with these facilities so coming back to the difference, when can you read their needs? one thing is that when the transaction is done and God forbid, there is always a risk of fraud so the fraud where you have to stop it, as soon as you have the data, you have to transform it or analyze it you cannot wait for it to be taken to the database first, then I will do something on it, then I will do a transformation and then I will take out its report so you have said that you can see this example from both the angles when you have the data, you have seen that this data, this particular transaction is fraudulent or this is okay after you see it immediately, you take it to your database and then you analyze it or the transaction which is fraudulent or in which you have a risk of fraud, then you segregate it and store it in your database for the further action, the relevant department then they analyze the data and see which bank account it is or which number it is from which such requests are coming so this is a very interesting subject, understanding of this step by step gradually will develop and in the fraud analytics, telecom, banking and in the basemar industry, it is a complete field for the data scientists it is very important to understand these things after ETL and then ELT, I have explained it in detail, but I think on the first slide, I have explained these things in a very good way but there is another topic which I would like to discuss with you, that is about data ingestion the concept of data ingestion is basically this has been evolving with the big data we had SQL data or CSV, that was a conventional data which was coming to us and we were analyzing it ETL was doing it, ELT was doing it, but the data ingestion is mostly this is with the real time in which you have the social media, IoT, or any other sources, you have the real time data and one of the interesting aspects is that the example I shared with you about banking fraud you can do that with data ingestion, it will depend on the data that you have done the data source for which you work, if it is still on the old technologies, then your data sources and data types are the same but if it is on the latest technology, then it is on the big data, then it is on the IoT I believe we have all experience fingerprints, retina recognition, so these are multiple types of data if you have to withdraw money from your ATM with your thumb impression, then it has to be a real time transaction what will happen is that if the database of the ATM application is no SQL database, then that data will be ingested and you will process it in the same way, if it is an old database, then it will go a bit like ETL or ELT but if it is a big data or any other tool or platform, then it is for that and the tools of the big data that we have discussed, when we were discussing big data like Kafka, Spark, your Scoop these are the tools that are the manipulation of the real time data or its comparison or reconciliation the real time we perform the services, then these are the tools that are used and the data ingestion is that you can use ETL and ELT in both the concepts that you can take the data as it is and analyze it and then you can keep the data in your database, data warehouse or data lake this is applicable everywhere, again it will depend on the situation of your source system these are the technologies that are your analytical platforms or the data that you have to use for self-service analytics or any other technique that you have to use to share your information or visualization these are the three concepts that can work in tandem with each other so this is about ETL, ELT and data ingestion