Pavan krishna teja

My name is. Uh, I have 6.5 years of experience in IT industry. So I have, um, uh, experience in front end and back end as well. So coming to front end, I have a good experience in React JS, HTML, CSS, JavaScript, uh, React, and Redux, TypeScript, uh, as well. And I have worked on micro front ends also. Uh, and, also, I have good experience in Python, Django So with post this, uh, I have worked on building few, uh, applications in Python and created the various SQL APIs and all. So yeah. Uh, over the past 6 years, I have been working in multiple, uh, technologies and created multiple projects and, uh, worked in, uh, various, uh, international clients like Standard Standard Bank and Lowe's and created various applications like, uh, insurance products and, uh, vendor compliance products and human resource management system and also, uh, work in a German based startup in which I have worked on, uh, GraphQL and Gatsby to build a, uh, blog website. Uh, so this is all about myself. Thank you.

So to build a, uh, React JS dashboard, uh, so I will use a usable components, basically. So in which, uh, the dashboard will be having various, uh, reasonable widgets and all. So, uh, I create separate components for each and every particular section. So each particular component will be utilized as a single component on the main screen. So, uh, yeah, in this way, I can create the customizable dashboard, uh, by using the React JS and Nexus as well. Uh, so, basically, for suppose if we are having some tables and also we are, uh, having some uh, data data part and all, and also we will be having some user information. So all these sections will be, uh, I will first divide all these sections into separate components. And then, uh, once each and every component is built separately, I will, uh, entirely include all these components into a single comp single screen. So where it meets, uh, I will name it as dashboard and it will be, uh, inside the main screen. So to, uh, enhance the enhance this dashboard screen with NextOS features, I will use NextOS, uh, file based routing. So in the Nextiva's, uh, I can create a folder, uh, in a particular I can create a file in inside a particular folder so that, uh, this part this file will be treated as this individual route. So if I name it as dashboard or JS, uh, it will, uh, directly, uh, come on the UI, uh, like, with the URL dashboard. And, uh, I can show the UI whatever inside the dashboard, uh, JS file as a, uh, dashboard page. So this is what I do, uh, when whenever whenever I need to clear some dashboard and all. Yeah. Thank you.

So, basically, next year, so So usually, next year, what happens is, uh, while the build time, uh, whatever the APIs are there, uh, for for a particular screen, uh, these APIs get called, uh, while the build time and, uh, all the data is, uh, entered into the application there itself. So based on that, our application, uh, UI is getting rendered, uh, in the server side itself. So we use matters like, uh, get get dynamic props, uh, in excess. So by utilizing this hook, we can make an API call by the build time and, uh, to we can create a static, uh, page. So in case of we are we are require we have we are having a requirement of generating dynamic content, uh, based on daily updates. We can use get get dynamic props, uh, hook so that, uh, we can fetch the data whenever required by using, uh, this, uh, method. So, yeah, I will do I will go ahead with this, uh, approach. Thank you.

Uh, so by utilizing the context API and hooks, uh, if I need to, uh, if I need to maintain the state across various, uh, components, uh, so I will go with this approach. Uh, like, basically, uh, with the context API, I'll create, uh, all the context required for state management, and then, uh, I will create the React hooks. So, basically, uh, this context will, uh, I will whichever I have created, I will wrap my context to, uh, onto the app, uh, main app dot j s file. So what are the values I'm I need to pass into the app? I can send the values to this to this wrapper. So once I wrap and I will send the values, I can get access to the values, uh, in all the components, which are the child components of the main main app component. So in this way, I can get all the required data from parent, uh, full parent component, uh, which is wrapped on inside a context wrapper. So yeah.

Okay. Uh, to manage session state one second. So if it is load balance, the React application with Django back end k. So, basically, I will go with, uh, set up in up a Django, like, uh, session management. So so in Django, we can enable session management. Uh, basically, Django provides us a framework, uh, we can in which we can save the session data, uh, on the server side. And, also, uh, users are cooking, uh, containing the session ID on the client side. Uh, and, also, we can have the authentication middleware for Django. So we need to ensure that, uh, Django's session and authentication middleware are enabled. And, uh, and also we need to take care of that, uh, Django's middleware, uh, is, uh, like, okay. Uh, so in this way, we can use Django's middleware, uh, to authenticate, uh, all the sessions. And, uh, and also we can use the Django rest framework for API endpoints, basically. So we need to install Django, uh, rest framework. And, uh, so this is used to build the API endpoints for handling authentication, uh, so such as login and logout. So we can use a a PIP install for this, uh, like, we can use the Django rest framework. Uh, like, this is a dependency. Yeah. And and then we need to create the endpoints, uh, for login and logout. And, also, we need to set up the Axios in React for API request. And so to handling the, uh, login and logout in React, uh, we can use the, uh, Axios, uh, third party library with the with the end point which we have created in the Django framework. And, also, we need to use the context, uh, for state management, uh, and session state. So I think and also we can, uh, predict the routes based on the second states. Um, so in this way, uh, it's a we can create a robust way to uh, manage our session state, uh, in React app by using a Django backend. Yeah.

Okay. Build and optimize the database. Okay. Build and optimize. So Yeah. To build an optimist MySQL schema. Okay. To build and optimize so we can follow, uh, key steps, uh, best basically, to ensure, uh, our database schema is well structured and, uh, it is normalized and optimized for performance. So, basically, Django provides, uh, high level ORM, uh, that allows you to define models, uh, which are then converted into SQL tables. Uh, so, basically, uh, we can do this, like, define, uh, we can define our models carefully. So we need to normalize data and then, uh, avoid over normalization. And also, we need to use appropriate field types. So, basically, we need to map our fields to appropriate Django fields. Example, like, uh, care field, integer field, and text field. And we need to set indexes and the unique constraints, uh, basically. So, like, we need to set the primary keys. So, basically, uh, by default itself, uh, Django adds primary key, uh, ID, uh, to each model. So we can, uh, specify different field as the primary keys if needed. Uh, so, basically, Django will keep the ID as a primary field, but we can specify different one. And also, we need to keep unique constraints, like, uh, add unique like, we need to add the unique, uh, constraint as true so that, um, uh, we can have the unique, uh, like, uniqueness in the fields. And, um, and also and also we can use Django's database options, uh, for performance improvement. So, basically, MySQL supports indexing options, uh, for improving query performance. So, for example, you can create composite indexes in Django's meta class, to optimize the multi column searches. And, also, we can use foreign key, uh, one to one field and, uh, um, and also many to many field, uh, so that we can get, uh, like, while defining the relationships, it will be very easy. And, uh, we can also optimize the query performance, uh, with the select related and the prefetch related. So, basically, select related is, uh, used for single value relationships, uh, such as foreign keys and all. Uh, and on the other hand, prefetch related is used for many to many and, uh, reverse foreign key relationships. Uh, so in this way, we can, uh, like, basically perform this upgrade lookup for each relationship and joins them in the Python.

Okay.

Yeah. So, yeah, basically, next year's okay. So, uh, incremental static regeneration, uh, basically, is used to improve the performance, uh, on on any of the applications, uh, using next year's. Uh, so, basically, it allows us to, uh, statically generate and then update, uh, like, the news content incrementally. Um, so this basically, it's a base very advantage of for next years. So, basically, the we can improve the performance, uh, in news portals. Uh, so by using ISR, uh, the the pages load very faster, and uh, the the content is served, uh, prerendered. So we don't need to fetch data at the time of load or we don't we don't need to create any HTML element while loading the screen. Instead of that, uh, our, uh, HTML content is rendered by, uh, like, even if even before the user is going to, uh, see the screen. So we can do this from uh, CDN, like, the content disputed networks and all so that, um, the loading speed is very higher. And, uh, and also we can get the news portals. Basically, news portals need the frequently updated data. So ISR allows this, uh, this feature by, uh, increment incrementally recentrating the data in the background, Like, after a set of interval, uh, the the data will get updated with the latest of fresh data without requiring, uh, to pull that data again and again. So in this way, we can optimize the performance of our application. And, also, by serving the prerendered pages, uh, I the ISR reduces the load on the server basically. So, uh, as it does not, uh, require on on demand server side rendering for each request. And, also, this allows, uh, new spotters to handle a large number of concurrent users because significant server costs. So yeah. Basically, to implement it, uh, we can, uh, define our pages, uh, in excels using get static props, uh, and also we can specify revalidate, uh, interval. Uh, with the help of get static props and revalid, we can ensure this. Yeah. Yeah. So, basically, this is what, uh, I approve I will go ahead with. Yeah.

Okay. So, basically, for cross browse cross browser, uh, compatibility of a particular CSS file, there are some libraries, uh, which we can install so that, uh, by installing these libraries, it will add the web kit or hyphen web kit, uh, hyphen onto our CSS files. So this particular feature will helpful for the cross compatible cross border compatible CSS and also for cross, uh, device compatible CSS, we can, uh, use, uh, media queries for different widths, uh, so that it will be compatible with all the devices, uh, in which the, uh, screen is getting loaded on.

Yeah. I'm not sure how to achieve this. So but I can think of this. So, basically, um, uh, I think, uh, we need to first try and save the model. So first, we need to try our mission learning model in Python, uh, using that list like Sky Skykit learn and TensorFlow or PyTorch. And once, uh, the train the model is trained, we need to save the model, uh, in a serialized format, um, such as Joblib or Pickle for Sky Skykit learn models or it's 5 for Keras, uh, or TensorFlow models. And, uh, so we need to load this model in Django. Uh, to do this, uh, we need to add the shared model file to our Django project. So, basically, within an appropriate directory, uh, example, like, we can create a new folder, uh, like models. And then, uh, we need to load the the model, uh, when the Django starts, Like, usually, uh, as a global variable. Uh, so it doesn't need to be loaded for each request. Uh, so you can load the model with the Django apps, uh, like app apps dot pi. Like, views are also utility functions can be used to load this, uh, Django model. And, also, need to create a prediction view. Uh, to do this, you need to define the Django view to handle, uh, the incoming request for predictions. So the view should accept the input data, uh, often via post request, uh, with the JSON payloads, and use the model to make predictions on the data. So we need to implement real time prediction logic as well. So, basically, uh, we can use this loaded model to make predictions on data received, process the data in the format expected by the model, and also invoke the model's prediction method and send the predictions as a response to the client. So, uh, in the last step, like, it's basically optional, uh, we can use Django as framework as well. So it's called BRF. Uh, basically, it simplifies, uh, simplifies handling JSON based API records and responses, uh, making it easier to set up your Django back end for API calls. Yeah. So this is basically how I do this.