Mar 26, 2026 Enterprise Brain Scrum - Transcript 00:00:00 Yeshwanth Reddy Yerraguntla: Hi bro. Good Rajashekar G: High number one. Yeshwanth Reddy Yerraguntla: morning. Kalakonda Harshith Rao: You're born in love. Yeshwanth Reddy Yerraguntla: I have s***. Satyasri Prabhakar Mantripragada: Hi all, good morning. Good evening. Kalakonda Harshith Rao: Good morning, sir. Rajashekar G: Good morning, sir. Yeshwanth Reddy Yerraguntla: I don't Amulya Maggidi: Put them on this side. Rajashekar G: Ablation Yeshwanth Reddy Yerraguntla: know. Rajashekar G: power joining. Yeshwanth Reddy Yerraguntla: 2 I'm back. Start them Rajashekar G: Yes sir. Now we can start. Yeshwanth Reddy Yerraguntla: up. Okay, I'm sharing my screen. So the agenda is uh I want to cover mainly in the code what are all the components that um are available so that you can start u building on those things. Um, let me take a screenshot and do some drawings. Sorry. So, most of you know that running make UI will u u launch the agents, right? Um Rajashekar G: Yes, summer. Yeshwanth Reddy Yerraguntla: right so the main code is present in chanaka folder um after this call I'll try to make some pushes but um you can see uh that there are four agents two data source agents one realtime agent and one u user aware agent So if I come here 00:07:17 Yeshwanth Reddy Yerraguntla: we have Gmail and Jira available. Both are working with mock data. Okay, we'll go into what is agent.py in a minute. Sorry, both are working with mock date. Both have agent.py. These are the only two things a docker file is needed for launching uh that one agent if you want it manually. Right? So the point of any agent um in whatever we are building is um in the worst case scenario that agent should work by itself even though other components are missing just if you want to test Jira agent you should be in a position to test only Jira. So if I if you launch makei it will show that you know there are these uh uh four uh ports running right so um 81 and 8002 are Jira and Gmail agents so 8002 / chat you have to type just in case I didn't use the root for any uh specific thing so you log in with Google. Okay. 00:08:47 Yeshwanth Reddy Yerraguntla: I don't think this is going to work. Yeah, sorry. So, you have to first launch with uh 8010. Uh actually, let me log out and log in. Now if I come and check 8,02 yeah Jira agent is available Gmail agent is available and at 83 user aware agent is available each agent will have its own uh dedicated u uh conversation histories. Ideally this is the conversation street should use and the mock data is present here. So for emails there is some mock data. For Jira there is some mock data. There is some mock data for users also. Who who are the users? What projects are they present in? What are their dashboard preferences? That kind of a dummy data is there. Snapshot is not a mock data. It is just uh it is collecting that mock data and it is um sort of building a snapshot based on demand. fake snapshot. 00:10:16 Yeshwanth Reddy Yerraguntla: So, uh we'll go into the details of what this does if time permits. Um but you can always ask charg what is happening here. It's just doing uh it is p pulling the entire um all the tickets and emails and uh for a specific date it is doing a filter. That's all it is doing. I think you can just read and figure it out. So that is the mark data. So coming and asking if I come here if I ask or even simply saying hi should work. So one uh interesting thing that I'm currently doing is uh I'm not using geminy flash um in our uh in this setup I'm using something locally I'm using some quen 3.5 model why because for the sake of debugging uh generally I feel we don't need to waste a lot of tokens and to use a small model will help will force us to think uh uh in a smart way. What I'm trying to say is if you use something very heavy naturally things will work. 00:11:40 Yeshwanth Reddy Yerraguntla: So you might write inefficient code and the LLM will still take the burden of making the uh whole pipeline work. But if you use a very small model usually what that means is um your code should be very efficient that it you should not send too much data to the LLM. um your LLM tool call should be pretty accurate, your instruction should be pretty accurate with these kinds of constraints. If you make your code work on quen 3.5 that basically means when you actually switch to geminy flash 3.0 it will anyway work there is a very good chance that it will work right. So uh this is one uh AI lesson that I also I think I learned over the last one week that try to make your pipeline work with a small model. Uh usually if it works that it means that you're writing very good code. So coming uh to now did I ask answer? Rajashekar G: uh and it's like uh I went and Yeshwanth Reddy Yerraguntla: Yeah. Hm. 00:12:51 Rajashekar G: see this hugging face thing. Yeshwanth Reddy Yerraguntla: H use Rajashekar G: So there will be commands to install it locally and run. Yeshwanth Reddy Yerraguntla: MLX. Rajashekar G: Yeah, MLX MLX community is there. Yeshwanth Reddy Yerraguntla: H correct. Correct. Rajashekar G: Uh which one like quen 3.5 397 bytes A7 B4 B is there different Yeshwanth Reddy Yerraguntla: Uh yeah, Rajashekar G: by Yeshwanth Reddy Yerraguntla: I mean use the 4bit version. Uh yeah. Rajashekar G: okay sure Yeshwanth Reddy Yerraguntla: I mean we it is it is good if we can go even smaller but what I observed is the smaller models in general community itself says they are very bad for general purpose programming like the cutff is at 9 billion parameters. So give it a try. Um cut anyway. So we have four agents. The main uh Chanaka is at 80010. You can also do some cheating here. So even though you logged in with your own Gmail and all that, you can actually come to admin uh here and uh fake uh a user. 00:14:05 Yeshwanth Reddy Yerraguntla: Right? So if I do Manisha and set and I come back uh in Chanaka if I ask now who am I it is sort of you know yeah it's saying you know you are Manisha and these are your preferences your team are currently unavailable in the system so this pipe plumbing still needs to take some effort But that setting of uh admin from 8010 local host 8010/admin once you do that that sort of reflects in the user aware agent also. So now if I ask uh new chat who am I what is my briefing today? It should say your Manisha and yeah whatever is happening it's happening from this local system your Manisha Gupta blah blah blah blah blah um relevant emails okay so what that basically means at a system level is um data agents to some extent is available. Realtime agent is available in the sense that it can you can ask uh questions to data agents and get back answers. Uh let me just uh confirm that that is actually the case. 00:15:49 Yeshwanth Reddy Yerraguntla: 8010 / chat. So it says Manisha what are my mails? Let us see. It should ideally ask uh Jira agent sorry Gmail agent and get back the answer. Ask Gmail what are based on the emails. Okay. Um how did it say no emails? Mark data emails. Manisha is CC from let me switch to Vijay Kini and ask the same question. Set. There are some emails, huh? Okay, it's actually okay. fetching from my actual Gmail. So I think this is a bug. What I mean is if you loging in it will fetch your emails also. It's not that my security is compromised or anything. Rajashekar G: Okay. Yeshwanth Reddy Yerraguntla: Um Rajashekar G: How actually mail connected like Gmail source implemented that Yeshwanth Reddy Yerraguntla: I thought it was using mock data. Rajashekar G: way. Yeshwanth Reddy Yerraguntla: Uh there were sorry back end chanaka Gmail agent. Oh it is using uh actual GWS services. 00:18:20 Yeshwanth Reddy Yerraguntla: Okay then uh I can ask RGP to use my let me do that dot backup. Um we use mock data from data /mark data. Okay, that will work. Uh but the point is uh the takeaway is um this pipeline is working right. It it can ask questions to agents and it can get back answers. Um if you ask something like uh generic, it might ask both the agents, right? So it might synthesize both of them and give an answer that is also there. So, um let me create a fresh layer. Gmail and Jira are working. This back and forth pipeline is working. And this is A2A uh enabled. And this is also has uh the uh metadata information like the ability to send which user is asking the question, the ability to um send if you are in training mode or um normal mode, ad hoc mode, if are are you in refresh mode, all these things are technically possible currently. 00:20:24 Yeshwanth Reddy Yerraguntla: Um I would say only this one is actually in the code. These two are pending. I mean we can always do them on uh when necessity comes right. It's just that you have to add a specific uh uh key in your JSON and it should work. What is also working is um now by the way um this is basically our mark data folder. So right now data agents are in a sense connected to mark data. So when you ask questions to data source agents they are fetching information from this only. User aware agent is also working. But whatever you are asking user aware agent is also giving you answers from mock data. It is just that uh here um it knows because which user is asking I will give them only those uh details user specific details. Right? So this is working. What is not currently present is this pipeline that you have a real-time intelligence agent that can update the organization snapshot based on the information from data agents. 00:21:51 Yeshwanth Reddy Yerraguntla: Uh that is missing right because we are still working with mock data. I use that as proxy for both data source agents and user aware agent. And like I said every agent that we have user agent data agents all of these are currently working on pylog interface there everyone every one of them has its own pylog interface and uh but this you know if you ask a question to a realtime agent from pylog interface only you can uh get this channel communication. and working. So whatever arrows I have drawn those are all the things that are available. Um is at least the highle picture clear? Rajashekar G: Yes sir. Guys, any questions please Yeshwanth Reddy Yerraguntla: Okay. Rajashekar G: ask. Efficient. Kalakonda Harshith Rao: How we replace mock data with real data like DB connections or something or any row creations Yes, I Yeshwanth Reddy Yerraguntla: How will we replace? Yeah, you have to use the actual DB connections. So, by accident, uh I showed right this agent.py.up. 00:23:13 Kalakonda Harshith Rao: understood Yeshwanth Reddy Yerraguntla: If you now take it, this is the actual re real agent source. Same thing with Jira also. There is one more agent part backup. This is actually connecting to your Jira and giving you information. So this is how you'll have to uh switch on the actual data sources. This did that answer the question? Kalakonda Harshith Rao: until unless real time data comes, we used to use the mock data which is present in the agent.py which is now you created mock data by. Yeshwanth Reddy Yerraguntla: Correct. Kalakonda Harshith Rao: So now if uh if we answer if we go to test that agent agent calls Yeshwanth Reddy Yerraguntla: Correct. Kalakonda Harshith Rao: with realtime data can we uh if if we change admin also we can't get the data of uh that admin Yeshwanth Reddy Yerraguntla: Oh, at that time you are now using only mark data. Admin is a mark admin. The user is a mark user. Kalakonda Harshith Rao: okay okay Yeshwanth Reddy Yerraguntla: So data will also be mark 00:24:26 Kalakonda Harshith Rao: so as uh the email agent which I wrote also in the makei file having the same separate file Yeshwanth Reddy Yerraguntla: data. Kalakonda Harshith Rao: which is like agent uh which goes for the uh different port. Rajashekar G: So actually what he's saying you know he he write he write an email sending module uh that's like that we can use it Kalakonda Harshith Rao: Yes. Yeshwanth Reddy Yerraguntla: Email sending module. Kalakonda Harshith Rao: Uh Rajashekar G: in future like to whenever we need to send an alert. Yeshwanth Reddy Yerraguntla: Uh for alert. Got it. Rajashekar G: So he write that as a micros service. So that is exposed on a port like let's say Yeshwanth Reddy Yerraguntla: Okay. Kalakonda Harshith Rao: 8020 is a port uh for that command to Rajashekar G: 8020. we can go there and uh colony 80080/ Kalakonda Harshith Rao: work. Rajashekar G: send email and uh hit that API with a payload that will send the mail. So that's a separate model. Yeshwanth Reddy Yerraguntla: Okay. Rajashekar G: Hashit right? Yes. Yeshwanth Reddy Yerraguntla: Got it. Rajashekar G: So what is your question here 00:25:22 Yeshwanth Reddy Yerraguntla: Um Kalakonda Harshith Rao: So if if I use that parallel service also the mock data will be considered to send the Rajashekar G: hash? Yeshwanth Reddy Yerraguntla: yeah Rajashekar G: Let's Yeshwanth Reddy Yerraguntla: man it comes down to how we implement it like to that to that Kalakonda Harshith Rao: mail. Yeshwanth Reddy Yerraguntla: uh service which is sending emails that service receives some email from someone right who that service has an entry point of to whom I should send the Kalakonda Harshith Rao: Yes. Yes. Yeshwanth Reddy Yerraguntla: email um there uh it depends on what we write. So if you write uh something like from the user aware agent send the information of who is the active user then to that user only it will send. Kalakonda Harshith Rao: What? Yeshwanth Reddy Yerraguntla: So it's a question of implementation. Kalakonda Harshith Rao: Shut. Yeshwanth Reddy Yerraguntla: So speaking of implementation we'll actually get into how these things are implemented because that is going to be critical uh in understanding. So the base uh um model is present in base.py. Maybe this will be changed into a different folder but just keep a look on base.py uh where the key class is called as enterprise agent. 00:26:32 Yeshwanth Reddy Yerraguntla: for the time being enterprise agent is um subclassing uh I don't know what is the right term enterprise agent's parent is u p identical a responder pdentic responder why we are using is because this remembers history and all those things and this acts well with pylog UI so enterprise agent may the I don't think there is anything uh new that you have to understand here it's just these are all wrappers on the actual agents uh functionalities tool tool plane you already know so these just these are just decorators um run is basically pidentic run right uh only thing I think it is doing trying to do is remember the message history during every run it will remember what is the message history is how the actual protocol works so when we are doing when we are launching make UI that the fact that all the agents are reserving their own um ports, right? And on those ports, who is listening is this A2A protocol is listening for new messages. Um so this is base.py. A2A context is where um we have the necessary plumbing to um uh send the user information. 00:28:03 Yeshwanth Reddy Yerraguntla: What I mean to say is um this kind of information right now when I am talking from a realtime agent to Jira agent realtime agent has to send some information along with the question about what is the context of the question. It could be who is asking the question. It could be which company, what kind of tenant is asking is the user a trainer or not a trainer or if there are any other uh uh keys that we are missing. Right? So you this request context is now creating all these uh uh keys for us. Um sorry it's not creating you just defining that this is how I am going to send between A2A agents. uh and during uh dependency uh okay one second let me just so when you call your core agent.py right when I open base.py by you are doing agent.r run. So uh agent agent worker I don't have exact spot where this thing gets called but essentially when your agent is doing agent.r run. 00:29:30 Yeshwanth Reddy Yerraguntla: You are uh internally attaching certain dependencies on the message uh packet that uh um here are here is the question going to be currently that question does not have any context. I will add my own context and then I will send the agent u my question and my context. That's what this is trying to do. Um again I think you can ask charg and get more I ideally I should be the one explaining this more properly. um maybe if we have some time I'll try to explore this more uh partial Vara Kumar Jagarapu: Will it take some time or I Yeshwanth Reddy Yerraguntla: for oh because we have a meeting I am there Vara Kumar Jagarapu: mean other call is there? Are you there or not? Yeshwanth Reddy Yerraguntla: actually I need to connect in that call also Vara Kumar Jagarapu: Otherwise, Yeshwanth Reddy Yerraguntla: Uh Vara Kumar Jagarapu: I will let them know like it will take five minutes or so. Yeshwanth Reddy Yerraguntla: fine. Wrap it in 10 minutes. Vara Kumar Jagarapu: Wait, sorry. 00:30:49 Yeshwanth Reddy Yerraguntla: Yeah, Vara Kumar Jagarapu: Uh, you're saying something I didn't get. Yeshwanth Reddy Yerraguntla: I'll wrap it in 10 minutes this call. Vara Kumar Jagarapu: Okay. Okay. Okay. Yeah. Yeshwanth Reddy Yerraguntla: So once we have that base.py Enterprise uh enterprise agent uh object every agent in our this thing now can use that. So here we have um where is it? We have Chanaka. Ideally these instructions is Chanaka's world model. Uh Chanaka will have its own set of skills. All those things will now come and attach here. Right? So instead of saying instructions, so ideally it should be like posgress connection uh string so that it can fetch the instructions from that database directly. Same thing with skills, same thing with uh runbooks and all those things that those are the things which need to change. But right now um yeah Chanaka is a normal agent uh which has A2A ability and uh there is a wrapper for asking downstream agents. 00:32:01 Yeshwanth Reddy Yerraguntla: Now all you need to do is take the URL ask your question and send your context. Context is again who is the user what all those things. So this is this entire function is just doing a a toa wrapping that you try to do message send then you wait on a for loop for uh so much time and get back the answer from the A2A agent all that. So how many other agents downstream data source agents need to be there all of them can be registered like this as a tool call. Um uh Rajar I just saw that chat. Rajashekar G: Yeah. Yeshwanth Reddy Yerraguntla: Um do you want me to address it now? Rajashekar G: And once you have completed at the end if you give us with Yeshwanth Reddy Yerraguntla: Okay sure. Rajashekar G: this we can start. Yeshwanth Reddy Yerraguntla: So I've done I mean one thing I have ensured is when you are creating a fresh agent you can always ask uh whatever copilot uh how to create a fresh agent or how to update it right so in diagram sorry uh where is it new agent guide in this you I have given everything about how to create a fresh agent um so that can also help you with the deployment aspect of that when you have an existing agent how do you deploy the docker 00:33:31 Yeshwanth Reddy Yerraguntla: uh commands are there the moment you do uh compose up Rajashekar G: Okay. Yeshwanth Reddy Yerraguntla: that one agent will go up so that that is how you can individually control each agent's deployment Rajashekar G: Even though RT agent RT agent need to deploy separately and uh one data source agent need to deploy separately like that. Yeshwanth Reddy Yerraguntla: cut Correct. Rajashekar G: Yeah. Yeshwanth Reddy Yerraguntla: Yeah. If you want all of them to be deployed in one go, Rajashekar G: And Okay. Yeshwanth Reddy Yerraguntla: for that also there is a docker compose. Rajashekar G: And one more thing we need to see is URLs like how they map the URLs with here we will be dealing with the port Yeshwanth Reddy Yerraguntla: So that is all config driven. Rajashekar G: side. Yeah, that Yeshwanth Reddy Yerraguntla: If I open uh uh so what is this Gmail agent Rajashekar G: to Yeshwanth Reddy Yerraguntla: port? Uh Gmail agent. Okay. Yeah. This is coming from this particular variable. Ideally, if you can put it in an environment file, it will work. 00:34:33 Yeshwanth Reddy Yerraguntla: Um, yeah, all of these are config driven. Rajashekar G: Yeah. Yeshwanth Reddy Yerraguntla: All the services are um properly um like first you have to put uh launch the down data source agents then only you'll have to launch anaka and then for janaka you can tell which agent is listening on which port. all those things for individual agents also you can give what is the p DB path and all those Rajashekar G: Okay. Yeshwanth Reddy Yerraguntla: things. So um should be relatively straightforward that when you want to launch or deploy a new agent it should not take too much effort. Rajashekar G: Okay, got Yeshwanth Reddy Yerraguntla: Um so Rajashekar G: it. Yeshwanth Reddy Yerraguntla: uh the 8way listens on the root folder. Rajashekar G: I Yeshwanth Reddy Yerraguntla: Py log listens on the / chat. These are the two main things. Um other than that I think user aware agent also is not doing anything different just that uh I didn't connect it yet. Like this I told you know this plumbing is missing. Other than that, 00:35:40 Rajashekar G: right Yeshwanth Reddy Yerraguntla: I don't think you even need to uh wait on that to start creating other Rajashekar G: here. I think uh one more point we are missing like uh if we convert this and connect with Yeshwanth Reddy Yerraguntla: things. Rajashekar G: the react front end right now it is with pilot. So one port 8001 is connected with uh uh some data source agent 80010 is connected with uh uh RT agent like that. If you connect with a single front end application react application there will be endpoints Yeshwanth Reddy Yerraguntla: See, Rajashekar G: only one endpoint we can write right one front end application. So, how can Yeshwanth Reddy Yerraguntla: you can always create duplicates and uh do that ideally. Rajashekar G: we Yeshwanth Reddy Yerraguntla: I see what you mean. Um see this this one guy, right? He is this guy is fetching from pantic responder. Um and uh if I open Rajashekar G: Yes. Yeshwanth Reddy Yerraguntla: shaky so here I am doing create core app for that agent. Okay. 00:36:48 Yeshwanth Reddy Yerraguntla: This is my pantic AI agent. Um this is the uh entire like this is this will just give me the full fast fast HTML application and that is mounting on / chat. So how to move this from this to react is basically um yeah I mean this this Rajashekar G: Thank you. Yeshwanth Reddy Yerraguntla: will still give you the necessary websocket endpoint so that other front ends can also talk to it. Rajashekar G: Yes, Yeshwanth Reddy Yerraguntla: Do you understand what I'm trying? Rajashekar G: if we if we need to do like here plog it's simple for the front end application react Yeshwanth Reddy Yerraguntla: No, no, wait one second. Rajashekar G: we need to do Yeshwanth Reddy Yerraguntla: I am saying the moment you create this create core app, Rajashekar G: right Yeshwanth Reddy Yerraguntla: it will give you a pylo endpoint. Well, and good. But it is also giving you in this one thing, Rajashekar G: Yes. Yeshwanth Reddy Yerraguntla: it is also giving you the websockets endpoint API charts. um where is websockets Rajashekar G: Okay. 00:37:51 Yeshwanth Reddy Yerraguntla: apps register websocket routes right so if I come here and check for my app uh appwath is Rajashekar G: Oops. Yeshwanth Reddy Yerraguntla: base path/ws so basically um we already have When I do SL I know this will not work obviously but if I do something like u network to low if I type hi this this U guy is actually calling a websocket endpoint which is Rajashekar G: Let's Yeshwanth Reddy Yerraguntla: um uh where is WS fail Rajashekar G: socket on top like there is all fetch Yeshwanth Reddy Yerraguntla: Reload. Rajashekar G: XHR here you can filter with socket. Right side. Same line. Right side on the top. Yeshwanth Reddy Yerraguntla: Right. Rajashekar G: Right side Yeshwanth Reddy Yerraguntla: Socket. Oh, okay. Uh, Rajashekar G: socket. Yeshwanth Reddy Yerraguntla: okay. This is missing the socket. Uh, interesting. See, I know that it uh exposes a websocket. Rajashekar G: Okay. Yeshwanth Reddy Yerraguntla: If you can catch hold of that one websocket location and uh attach it to your uh uh React agent, that should be sufficient, 00:39:32 Rajashekar G: Okay, Yeshwanth Reddy Yerraguntla: right? So, you don't need to really invent uh uh again. Rajashekar G: now we need to have that one endpoint contrib with the front end Yeshwanth Reddy Yerraguntla: Correct. Rajashekar G: application. Yeshwanth Reddy Yerraguntla: So I mean I'm sure we can do an audit of pylog and get a strategy. Um but yeah most of the hard part we have already done just that the uh react endpoint should uh be Rajashekar G: Okay. Yeshwanth Reddy Yerraguntla: compatible. So the branch we need to start from is again chanaka/alpha. Uh we are running out of time. Rajashekar G: Okay. And Yeshwanth Reddy Yerraguntla: Deployment I already gave you env file for the latest uh Rajashekar G: And chisha/alpha is hardcoded with quen model right Yeshwanth Reddy Yerraguntla: um no it is not because uh if you see Rajashekar G: now. Yeshwanth Reddy Yerraguntla: core.py sorry if you see base.py by huh Rajashekar G: Oh yeah. Mix chart model is a class. Yeshwanth Reddy Yerraguntla: it will work only if you have MLX model in one of your environment variables if this is missing it will skip and come to light LLM yeah 00:40:46 Rajashekar G: Okay, got it. Yeah, Yeshwanth Reddy Yerraguntla: um so we'll drop off and connect back in 1 hour probably Rajashekar G: sure. Yeshwanth Reddy Yerraguntla: you can set up another call whatever convenience people have yeah harsh Rajashekar G: Sure. Yeshwanth Reddy Yerraguntla: s***. Phawhan Saii Gajjalakonda: Oh, Yeshwanth Reddy Yerraguntla: Oh, Phawhan Saii Gajjalakonda: actually Yeshwanth Reddy Yerraguntla: no one Phawhan Saii Gajjalakonda: like after listening. Rajashekar G: Your voice is breaking. Your voice is breaking. Yeshwanth Reddy Yerraguntla: pray. Phawhan Saii Gajjalakonda: Okay. Rajashekar G: Yeah, it's clear now. Phawhan Saii Gajjalakonda: Uh hello high level. Rajashekar G: Oh, Phawhan Saii Gajjalakonda: This is my second which I have attended. Rajashekar G: one. Yeshwanth Reddy Yerraguntla: Oh. Phawhan Saii Gajjalakonda: So like I was listening. Yeshwanth Reddy Yerraguntla: Oh no. Phawhan Saii Gajjalakonda: So I wanted to get know about these please like so that like if somebody can Yeshwanth Reddy Yerraguntla: H Phawhan Saii Gajjalakonda: explain if you give me the documentation Yeshwanth Reddy Yerraguntla: everything is there in this folder. Phawhan Saii Gajjalakonda: available. Yeshwanth Reddy Yerraguntla: Okay. Whatever every day we are discussing is in this one 00:42:14 Phawhan Saii Gajjalakonda: So uh uh yes and then one uh like like uh to whatever Yeshwanth Reddy Yerraguntla: folder. Phawhan Saii Gajjalakonda: I want to do next. Do we have do we have uh documentations already done for the Yeshwanth Reddy Yerraguntla: Yes. Phawhan Saii Gajjalakonda: requirements? Yeshwanth Reddy Yerraguntla: To whatever extent it is there, it's available. Rajashekar G: So power Phawhan Saii Gajjalakonda: and like uh who is the right guy Rajashekar G: our Phawhan Saii Gajjalakonda: uh where I can go and ask for requirements and he and he is responsible to give me all the requirements so that I can do my job and without deviating actually okay thank Rajashekar G: I I will Yeah know if I need anything Phawhan Saii Gajjalakonda: you Rajashekar G: from anyone I'll link up with whenever needed. Phawhan Saii Gajjalakonda: thank Actually Yeshwanth Reddy Yerraguntla: Okay. Rajashekar G: Yeah. Yeshwanth Reddy Yerraguntla: Now Phawhan Saii Gajjalakonda: last Yeshwanth Reddy Yerraguntla: sir sir I will drop a check or talk with everyone get their timings set Rajashekar G: Yeah, Yeshwanth Reddy Yerraguntla: up one more. Phawhan Saii Gajjalakonda: Uh Yeshwanth Reddy Yerraguntla: Bye. Rajashekar G: sure. 00:43:29 Rajashekar G: Okay. Okay. Phawhan Saii Gajjalakonda: office. Rajashekar G: Okay. Phawhan Saii Gajjalakonda: Oh, okay. Rajashekar G: So the point is uh guys sorry once we are in office uh go to chanika/alpha I asked for env if you share that he will share and what we will do we know right uh we'll initiate uv in the root folder we will run the make u command is make u if we run that the agents will be started running right uh we have tried that right A hash we have already Kalakonda Harshith Rao: Now make your Abhilash Adunuri: Main character. Rajashekar G: tried okay let's sit and try today and pray that first Kalakonda Harshith Rao: way. Rajashekar G: thing previous we need to close your jas by 11 a.m. first thing and once that is done uh pavan I have shared you a vasa link and also the a uh drive which is having all the excel files. So what we need to do is by today end of the day you need to go through that Vasa file and understand the data structures we have and already Ashwantana is saying that is very he's very confident that all the DB schema can be provided using that Vasa file directly. 00:45:05 Rajashekar G: So based on that we need to dump that entire data into a database SQL database and uh Abby or Amulia or Amulia will be working on the UI part. Amia you explore this uh new repository and how we need to work with react application to integrate this. Harshett you need to get Harshett you need to get the KT from Ganesh how that Kalakonda Harshith Rao: Okay. Now based on exponentially Rajashekar G: uh dashboards uh that package how the package is working dashboards package and Kalakonda Harshith Rao: okay we will set up a Rajashekar G: how we can integrate that with our our system into prior that first Kalakonda Harshith Rao: uh small meeting call with organization. Rajashekar G: make sure you guys have have the hands on with the current Kalakonda Harshith Rao: Yeah. Rajashekar G: repository Kalakonda Harshith Rao: I have current uh hands on the repository as I made the micros service. No problem. I know something Rajashekar G: Okay. And Kalakonda Harshith Rao: here. Rajashekar G: related sit and make sure something is moving. Okay. 00:46:12 Phawhan Saii Gajjalakonda: Before that I wanted to go through the drives or links whatever you shared we are selling and all after set up a Rajashekar G: Um drive based on the drive only he created this vasa link. Phawhan Saii Gajjalakonda: call okay Rajashekar G: So uh that should be done by before Phawhan Saii Gajjalakonda: I'll get to understand those and then I'll set up a call if require Rajashekar G: lunch time most ideally is expecting that is expecting by end of the day at least we need to connect the data to this thing and uh see something is running if you ask questions it should answer in Phawhan Saii Gajjalakonda: Okay, Rajashekar G: pilot Phawhan Saii Gajjalakonda: let let me check whether I can complete it by uh before lunch or not. Rajashekar G: okay and Shall I set up the next call at 11:00? Kalakonda Harshith Rao: Fine. Rajashekar G: Right. Phawhan Saii Gajjalakonda: Oh Rajashekar G: Right. Kalakonda Harshith Rao: Fine. Rajashekar G: But there will be scrum at 10:00 a.m. Who will be attending? Who will be not attending? Kalakonda Harshith Rao: Oh s***. We will join exactly at that 00:47:25 Rajashekar G: Scrum Kalakonda Harshith Rao: time. Amulya Maggidi: Yes Rajashekar G: start Phawhan Saii Gajjalakonda: yes. Amulya Maggidi: ma'am. Rajashekar G: everyone everyone come to office I'll tell Palash Phawhan Saii Gajjalakonda: Obviously. Rajashekar G: uh to make this come when we are in office or on the same call come to office like as early as Kalakonda Harshith Rao: So, Amulya Maggidi: Okay. Rajashekar G: possible let's have I I will put a call at Kalakonda Harshith Rao: so Rajashekar G: 11:00 Kalakonda Harshith Rao: where am I? Rajashekar G: okay Amulya Maggidi: different. Rajashekar G: We need to see how it works. Amulya Maggidi: Exactly. Rajashekar G: And what I'm saying is in pylock for our development version we need to have multiple data sources uh multiple endpoints but whenever we need to work on the real Amulya Maggidi: until Rajashekar G: platform single react application that can be linked with RT agent is enough Amulya Maggidi: multiple Rajashekar G: color RT agent is enough color why we need to have multiple data sources again to have it Amulya Maggidi: ports. Rajashekar G: in the Amulya Maggidi: Is this someone agent? Rajashekar G: uh we will be communicating with the RT agent. Our front end application will be communicating with RT agent in the back end. Amulya Maggidi: Okay. Rajashekar G: The data source will be enabled or not that back end will figure it out. Amulya Maggidi: Okay. Rajashekar G: No need to have the dependent front end. Amulya Maggidi: Well, Rajashekar G: What I'm thinking right now is there should be a flag basic system in local. Amulya Maggidi: okay. Rajashekar G: If I enable one flag that should run with the py if I enable something else that should link with the react likewise we need to see why because our development will will become speed Amulya Maggidi: Okay. Rajashekar G: right that we'll see. Amulya Maggidi: Yes. Okay. Okay. Transcription ended after 00:50:00 This editable transcript was computer generated and might contain errors. People can also change the text after it was created.