WEBVTT 00:00:05.219 --> 00:00:07.339 Welcome to Data and AI with Mukundan, where you 00:00:07.339 --> 00:00:10.400 learn AI by building. Hey, I'm Mukundan. This 00:00:10.400 --> 00:00:13.240 show is about solving real problems with small, 00:00:13.279 --> 00:00:16.019 useful AI, things that you can actually use today. 00:00:16.760 --> 00:00:20.539 Each week, we pick one problem, build a simple 00:00:20.539 --> 00:00:22.600 workflow or tool, and talk through the decisions, 00:00:22.839 --> 00:00:26.179 what to automate, how to check quality, and how 00:00:26.179 --> 00:00:30.300 to make it reliable. If you're a builder, creator, 00:00:30.500 --> 00:00:33.950 or just AI curious, You leave with the steps 00:00:33.950 --> 00:00:41.289 that you can copy tonight. You know that feeling 00:00:41.289 --> 00:00:45.030 when your calendar is a mess, your inbox is overflowing, 00:00:45.090 --> 00:00:49.509 and you think, if only I had an assistant who 00:00:49.509 --> 00:00:53.789 just handled it. As the promise of AI agents, 00:00:54.170 --> 00:00:57.469 last episode we defined what an agent really 00:00:57.469 --> 00:01:03.420 is. Today, we're going deeper. I'm going to show 00:01:03.420 --> 00:01:07.019 you how agents actually think. The three ingredients 00:01:07.019 --> 00:01:10.120 that make them useful is planning, memory and 00:01:10.120 --> 00:01:13.719 feedback loops. By the end, you'll see how to 00:01:13.719 --> 00:01:17.159 build an agent that doesn't just talk. It takes 00:01:17.159 --> 00:01:20.799 action, learns from mistakes and saves you real 00:01:20.799 --> 00:01:24.799 hours in your week. Here's the truth. Most agents 00:01:24.799 --> 00:01:28.319 you see on social media are toys. They look flashy 00:01:28.319 --> 00:01:32.709 in a demo but fall apart in real life. Why? Because 00:01:32.709 --> 00:01:36.689 they can't plan, they can't remember, and they 00:01:36.689 --> 00:01:40.469 can't fix mistakes. Without those three, an agent 00:01:40.469 --> 00:01:44.310 is just a chatbot with a fancier name. But when 00:01:44.310 --> 00:01:47.609 you add planning, memory, and feedback, something 00:01:47.609 --> 00:01:51.049 magical happens. Feels like the agent actually 00:01:51.049 --> 00:01:54.390 thinks. That's what we'll cover today with stories, 00:01:54.510 --> 00:01:56.849 a live demo, and even a quiz to test your instincts. 00:01:57.069 --> 00:02:00.150 Let's look at planning. Think of it, think of 00:02:00.150 --> 00:02:03.459 planning as the map. An agent starts with a goal, 00:02:03.560 --> 00:02:06.859 then breaks it to steps. Say the goal is plan 00:02:06.859 --> 00:02:10.319 my workout week. A chatbot might just list some 00:02:10.319 --> 00:02:14.280 exercises. But an agent, it checks your calendar, 00:02:14.560 --> 00:02:18.560 sees that you're free Monday morning or Monday 00:02:18.560 --> 00:02:23.139 evening and Wednesday morning. It balances intensity 00:02:23.139 --> 00:02:27.719 so that you're not sore two days in a row. And 00:02:27.719 --> 00:02:32.030 also it proposes a full week plan. That's planning. 00:02:32.310 --> 00:02:35.569 It takes a big task and turns it into manageable 00:02:35.569 --> 00:02:39.729 steps. Without it, you're wandering without a 00:02:39.729 --> 00:02:43.030 map. Let's look at memory. Now planning is great, 00:02:43.169 --> 00:02:46.430 but if you forget what you just did, you're stuck. 00:02:47.550 --> 00:02:51.930 That's why memory matters. Agents have two kinds 00:02:51.930 --> 00:02:55.530 of memory. Short -term memory is like a scratch 00:02:55.530 --> 00:02:59.599 pad. It tracks the last step or two. Now, long 00:02:59.599 --> 00:03:01.780 term memory for agents, it's like a notebook. 00:03:02.840 --> 00:03:06.919 It stores important things for later. For example, 00:03:06.919 --> 00:03:09.819 when scheduling your workouts, the agent remembers 00:03:09.819 --> 00:03:12.780 it already filled Monday. That's short term. 00:03:13.639 --> 00:03:16.379 It also remembers that you like running outdoors. 00:03:16.879 --> 00:03:22.419 So it avoids rainy days. That's long term. Without 00:03:22.419 --> 00:03:26.319 memory, every step is a fresh start. With memory, 00:03:26.400 --> 00:03:29.370 the agent feels consistent. Almost human -like. 00:03:29.490 --> 00:03:32.550 And here's the final ingredient of agents. Feedback 00:03:32.550 --> 00:03:35.490 loops. Now when every plan meets reality, reality 00:03:35.490 --> 00:03:38.330 pushes back. Let's say the agent tries to book 00:03:38.330 --> 00:03:41.590 Tuesday at 7pm but finds that the time is already 00:03:41.590 --> 00:03:45.550 taken. With feedback, it retries Wednesday morning. 00:03:46.490 --> 00:03:48.849 Or maybe it drafts an email that's too long. 00:03:48.949 --> 00:03:52.349 With feedback, it shortens it. Feedback can come 00:03:52.349 --> 00:03:55.349 from the system itself like automatic retries 00:03:55.349 --> 00:03:59.680 or from you, the human. which would be approval 00:03:59.680 --> 00:04:03.879 steps, yes or no buttons. Without feedback, the 00:04:03.879 --> 00:04:07.120 agents collapse at the first mistake. With feedback, 00:04:07.219 --> 00:04:09.180 they course correct and actually get things done. 00:04:10.000 --> 00:04:12.680 Okay, now let's look at a demo of the AI agent 00:04:12.680 --> 00:04:17.360 so that we can see how it actually works. Here's 00:04:17.360 --> 00:04:20.339 what happened. I first gave it a prompt and this 00:04:20.339 --> 00:04:24.199 was my prompt. I said, you are a calendar coach 00:04:24.199 --> 00:04:29.740 in agent mode. 90 -minute deep work sessions 00:04:29.740 --> 00:04:32.839 this week, plus one 45 -minute admin block and 00:04:32.839 --> 00:04:36.779 one 30 -minute workout. Constraints, week, window, 00:04:36.920 --> 00:04:42.560 Sunday, September 21 to September 27, 2025. And 00:04:42.560 --> 00:04:45.379 I mentioned the time zone admin. I also mentioned 00:04:45.379 --> 00:04:48.699 avoid conflicts with events I provide that treat 00:04:48.699 --> 00:04:52.339 as read -only ground truth. Prefer morning for 00:04:52.339 --> 00:04:55.079 deep work, admin after lunch, and all blocks 00:04:55.079 --> 00:04:58.920 end by 5 p .m. Ask for explicit approval. Do 00:04:58.920 --> 00:05:02.779 not create or modify events. Dry run only. If 00:05:02.779 --> 00:05:05.480 a proposed time overlaps, use a feedback loop 00:05:05.480 --> 00:05:07.759 to retry the next best slot. My events for this 00:05:07.759 --> 00:05:11.060 week. I mentioned what those events are. And 00:05:11.060 --> 00:05:13.839 I also gave what an output I'm expecting. I mentioned 00:05:13.839 --> 00:05:16.259 the output is around the proposed schedule. There 00:05:16.259 --> 00:05:19.199 is ISO local times. Each time with a one -line 00:05:19.199 --> 00:05:22.079 reason. And I also mentioned I want a yes -no 00:05:22.079 --> 00:05:24.800 approval checklist for each block. A two -sentence 00:05:24.800 --> 00:05:27.790 conflict rationale. even if there are no conflicts 00:05:27.790 --> 00:05:30.870 and a 60 second kickoff action for the first 00:05:30.870 --> 00:05:35.629 deep work block so i also gave it my actual calendar 00:05:35.629 --> 00:05:39.170 screenshot so what it did was attribute this 00:05:39.170 --> 00:05:42.350 is what it gave me back right it said i reviewed 00:05:42.350 --> 00:05:45.350 the calendar screenshot you shared the existing 00:05:45.350 --> 00:05:48.110 recurring events morning affirmations whatever 00:05:48.110 --> 00:05:51.329 i had you just mentioned the different events 00:05:51.329 --> 00:05:55.959 i have And they also mentioned that it does not 00:05:55.959 --> 00:05:58.500 conflict with the proposed blocks, which the 00:05:58.500 --> 00:06:00.759 AI was suggesting, the agent was suggesting. 00:06:01.579 --> 00:06:04.899 It also mentioned that the AI agent also mentioned 00:06:04.899 --> 00:06:08.319 that your weekends also have brief blocks at 00:06:08.319 --> 00:06:11.819 certain times and another brief block at certain 00:06:11.819 --> 00:06:13.540 time, but these are outside the proposed times. 00:06:13.680 --> 00:06:16.120 So the proposed schedule that it came up with, 00:06:16.139 --> 00:06:20.639 with no conflicts detected, had like four columns 00:06:20.639 --> 00:06:24.040 to it. One was a block. one was a proposed time 00:06:24.040 --> 00:06:28.240 one was reason and one was approve yes or no 00:06:28.240 --> 00:06:32.100 so one was like could be event one and between 00:06:32.100 --> 00:06:37.699 9 a .m to 10 30 a .m on 9 22 so on monday and 00:06:37.699 --> 00:06:40.160 then reason monday morning start allows for focused 00:06:40.160 --> 00:06:42.339 work while energy is high and then it gave me 00:06:42.339 --> 00:06:45.459 an approve yes or no with the checkbooks and 00:06:45.459 --> 00:06:47.459 then like four other events like that with an 00:06:47.459 --> 00:06:50.300 admin block uh event as well said that reason 00:06:50.300 --> 00:06:54.519 for admin block between 1 a 1 pm to 1 45 pm is 00:06:54.519 --> 00:06:56.100 because post lunch session for administrative 00:06:56.100 --> 00:06:59.220 tasks when cognitive load is lower so it's yes 00:06:59.220 --> 00:07:02.379 or no so here's what i'm noticing right it's 00:07:02.379 --> 00:07:07.959 taking into account what i wanted i said i wanted 00:07:07.959 --> 00:07:11.040 to do admin work after lunch focused work in 00:07:11.040 --> 00:07:15.879 the morning evening time for the workout And 00:07:15.879 --> 00:07:17.759 it gave me that and it gave me those proposed 00:07:17.759 --> 00:07:20.800 times and it said, this is what I think you should 00:07:20.800 --> 00:07:23.319 do and give me an approval so that I can go ahead 00:07:23.319 --> 00:07:25.420 and make the schedule. And it also mentioned 00:07:25.420 --> 00:07:27.100 like a kickoff action for the first deep work 00:07:27.100 --> 00:07:29.680 block. It said before starting the Monday deep 00:07:29.680 --> 00:07:32.459 work session, clear your workspace, silence notifications, 00:07:32.779 --> 00:07:34.899 and then write down the single most important 00:07:34.899 --> 00:07:37.560 task you intend to complete. Set a 90 minute 00:07:37.560 --> 00:07:39.819 timer, take a few deep breaths and jump straight 00:07:39.819 --> 00:07:43.079 into that task without checking messages or emails. 00:07:43.709 --> 00:07:45.490 Let me know which blocks you approve and I'll 00:07:45.490 --> 00:07:47.930 prepare the calendar events accordingly. Nothing 00:07:47.930 --> 00:07:50.209 will be added until you confirm. This is what 00:07:50.209 --> 00:07:53.810 the AI agent said to me. And here's the thing, 00:07:53.829 --> 00:07:58.170 right? I didn't mention like I work from 5 a 00:07:58.170 --> 00:08:03.230 .m. to 2 p .m. This kind of information would 00:08:03.230 --> 00:08:05.329 have been helpful for it. And it's obviously 00:08:05.329 --> 00:08:07.810 like I can obviously go back and do that, but 00:08:07.810 --> 00:08:12.089 I just wanted to do like a general demo walkthrough 00:08:12.089 --> 00:08:16.480 here. help you see what how an agent really works 00:08:16.480 --> 00:08:20.379 what we need to see what we can see now is that 00:08:20.379 --> 00:08:23.240 ai agents can help you prepare your calendar 00:08:23.240 --> 00:08:27.060 if that was already not known it's just something 00:08:27.060 --> 00:08:31.480 that is taken care of for you via you know without 00:08:31.480 --> 00:08:35.340 if you just notice the live demo here's what 00:08:35.340 --> 00:08:38.740 just happened i wanted to showcase the planning 00:08:38.740 --> 00:08:41.970 memory and feedback loop In terms of planning, 00:08:42.110 --> 00:08:44.970 it gave me a deep work session, which I wanted 00:08:44.970 --> 00:08:48.230 to do across Monday, Wednesday, Friday with a 00:08:48.230 --> 00:08:51.230 slaughtered admin in the afternoon and a workout 00:08:51.230 --> 00:08:53.509 in the evening. Memory, it remembered the window 00:08:53.509 --> 00:08:56.090 that I wanted to do it between, like it said, 00:08:56.190 --> 00:08:59.590 like September 27 was the end date, 21 being 00:08:59.590 --> 00:09:02.409 the start date. It remembered the window that 00:09:02.409 --> 00:09:04.509 I gave it and it didn't repeat any overlapping 00:09:04.509 --> 00:09:08.409 times. And in terms of feedback loop, when I 00:09:08.409 --> 00:09:11.470 added my calendar screenshot, it checked against 00:09:11.470 --> 00:09:14.889 those affirmations plus focus hours and explained 00:09:14.889 --> 00:09:17.230 why there were no conflicts and the humor in 00:09:17.230 --> 00:09:19.909 the loop it gave me an approval checklist whether 00:09:19.909 --> 00:09:22.590 yes or no whether i approve it or not showing 00:09:22.590 --> 00:09:26.289 the dry run board and a kickoff action like it 00:09:26.289 --> 00:09:29.509 ended with a concrete 60 second kickoff which 00:09:29.509 --> 00:09:32.649 was essentially that you know before starting 00:09:32.649 --> 00:09:36.389 the work on monday uh for deep work clear your 00:09:36.860 --> 00:09:39.019 workspace silence notifications and then write 00:09:39.019 --> 00:09:41.600 down single most important task you intend to 00:09:41.600 --> 00:09:45.059 complete so that is like a that's like a great 00:09:45.059 --> 00:09:49.539 kickoff of 60 seconds that that helps you to 00:09:49.539 --> 00:09:52.100 you know start your work right away and this 00:09:52.100 --> 00:09:54.500 whole thing is this is just gold for how ai agents 00:09:54.500 --> 00:09:57.080 actually think now i'm going to come back to 00:09:57.080 --> 00:09:59.320 the segment just want to talk about some other 00:09:59.320 --> 00:10:04.019 examples which where you can use ai agents like 00:10:04.019 --> 00:10:06.700 a calendar blocker For example, which I just 00:10:06.700 --> 00:10:08.419 talked about, where it reads your goals, scans 00:10:08.419 --> 00:10:10.700 your schedule, proposes blocks, and asks for 00:10:10.700 --> 00:10:14.840 approval before creating them. And email triage, 00:10:14.840 --> 00:10:17.399 which I mentioned last time as well. It sorts 00:10:17.399 --> 00:10:19.840 your emails by type, drafts short replies, and 00:10:19.840 --> 00:10:21.860 leaves the final sent to you. So the human in 00:10:21.860 --> 00:10:24.600 the loop is also there. And a podcast prep agent 00:10:24.600 --> 00:10:28.360 gathers bios, writes a briefing, and drafts starter 00:10:28.360 --> 00:10:30.940 questions. This could be for your podcast guest 00:10:30.940 --> 00:10:34.049 if you want to. have them and an inbox to notion 00:10:34.049 --> 00:10:36.629 connector so it turns your emails into to do 00:10:36.629 --> 00:10:39.330 tasks with due dates right so that that also 00:10:39.330 --> 00:10:42.610 is a great use case for an ai agent you can also 00:10:42.610 --> 00:10:45.830 have a data digest agent where it pulls yesterday's 00:10:45.830 --> 00:10:49.649 metrics flags anything unusual and drafts a slack 00:10:49.649 --> 00:10:52.409 update or a team update microsoft teams none 00:10:52.409 --> 00:10:55.190 of these are science fiction right there's practical 00:10:55.190 --> 00:11:00.399 helpers once you can use how And once you can 00:11:00.399 --> 00:11:02.179 know how to use one, you'll wonder how you ever 00:11:02.179 --> 00:11:03.960 worked without it. Okay, now that we learned 00:11:03.960 --> 00:11:07.720 AI agents, before I wrap up, I just wanted to 00:11:07.720 --> 00:11:12.100 do a true -false lightning round and answers 00:11:12.100 --> 00:11:14.720 at the end. All right, here's the first question. 00:11:15.259 --> 00:11:18.019 An AI agent is just a chatbot that answers questions, 00:11:18.179 --> 00:11:21.500 true or false. An AI agent is just a chatbot 00:11:21.500 --> 00:11:24.730 that answers questions. If an agent can replan 00:11:24.730 --> 00:11:27.070 when one tool breaks, it's probably real. The 00:11:27.070 --> 00:11:28.909 third question I have is, clear goals matter 00:11:28.909 --> 00:11:31.070 less than clever prompts. Again, all of these 00:11:31.070 --> 00:11:34.529 are true or false. Third, repeating again, clear 00:11:34.529 --> 00:11:37.570 goals matter less than clever prompts. Fourth 00:11:37.570 --> 00:11:42.110 one, memory makes agents feel consistent. Memory 00:11:42.110 --> 00:11:46.509 makes agents feel consistent. Fifth one, feedback 00:11:46.509 --> 00:11:50.169 loops are optional. Fifth one, again repeating, 00:11:50.409 --> 00:11:55.080 feedback loops are optional. Sixth one is agents 00:11:55.080 --> 00:11:58.019 don't have to be perfect. They just need to keep 00:11:58.019 --> 00:12:00.600 going. Repeating it again. Agents don't have 00:12:00.600 --> 00:12:02.899 to be perfect. They just need to keep going. 00:12:03.860 --> 00:12:07.059 All right. Now that I went with the questions, 00:12:07.120 --> 00:12:11.220 I just want to give you the answers here. The 00:12:11.220 --> 00:12:15.659 first answer here is false to the question. An 00:12:15.659 --> 00:12:17.860 AI agent is just a chatbot that answers questions. 00:12:18.200 --> 00:12:20.580 So that one is false. The second one question 00:12:20.580 --> 00:12:24.399 I'll repeat. And I'll give you the answer. If 00:12:24.399 --> 00:12:27.220 an agent can replan when one tool breaks, it's 00:12:27.220 --> 00:12:29.860 probably real. And the answer here is true. The 00:12:29.860 --> 00:12:32.580 third question. Clear goals matter less than 00:12:32.580 --> 00:12:35.159 clever prompts. The answer is false. Fourth one. 00:12:35.220 --> 00:12:39.019 Memory makes agents feel consistent. Answer is 00:12:39.019 --> 00:12:40.879 true. Fifth one. Feedback loops are optional. 00:12:41.779 --> 00:12:44.159 Answer is false. Sixth. Agents don't have to 00:12:44.159 --> 00:12:47.600 be perfect. They just need to keep going. Answer 00:12:47.600 --> 00:12:50.879 true. Well, if you scored all six, you're ahead 00:12:50.879 --> 00:12:53.090 of the curve. If not, don't worry, just learn 00:12:53.090 --> 00:12:56.090 by building. And the explanations for these answers 00:12:56.090 --> 00:12:58.629 are there in the episode. So just go back and 00:12:58.629 --> 00:13:01.450 listen to the episode and you'll figure out these 00:13:01.450 --> 00:13:05.610 answers. How this applies to your work. So whether 00:13:05.610 --> 00:13:07.690 you're studying, managing projects, or just battling 00:13:07.690 --> 00:13:10.409 your inbox, the same three ingredients apply. 00:13:10.970 --> 00:13:13.149 Where you need to define a clear goal, break 00:13:13.149 --> 00:13:15.970 it into steps, keep notes on what's been done, 00:13:16.110 --> 00:13:18.909 and add a way to adjust when things go wrong. 00:13:19.350 --> 00:13:21.929 That's it. That's the recipe. planning memory 00:13:21.929 --> 00:13:25.409 and feedback or if you know if you ignore these 00:13:25.409 --> 00:13:29.049 you're building sand castles if you include them 00:13:29.049 --> 00:13:32.289 you're building systems that last here's my challenge 00:13:32.289 --> 00:13:36.070 for you build your first mini agent i want you 00:13:36.070 --> 00:13:38.570 to pick one goal this week write it down in plain 00:13:38.570 --> 00:13:41.809 english or whatever language that you want break 00:13:41.809 --> 00:13:46.470 it into three steps add a scratch pad a dock 00:13:46.470 --> 00:13:49.889 a note or a sticky and add one feedback rule 00:13:49.889 --> 00:13:53.450 like Ask me before saving. Congratulations. That's 00:13:53.450 --> 00:13:56.690 your first mini agent. It won't be perfect, but 00:13:56.690 --> 00:13:59.129 it will work better than 90 % of the hype you 00:13:59.129 --> 00:14:00.950 see online. If your agent doesn't give you time 00:14:00.950 --> 00:14:03.830 back, it's just a toy. But when you frame the 00:14:03.830 --> 00:14:07.330 goal clearly, give it memory and let it learn 00:14:07.330 --> 00:14:10.090 from feedback. It becomes a tool you'd rely on 00:14:10.090 --> 00:14:14.710 every day. For me, the difference was real. I 00:14:14.710 --> 00:14:16.909 went from staring at a messy inbox and calendar 00:14:17.580 --> 00:14:19.960 to getting two hours back every week. That's 00:14:19.960 --> 00:14:22.100 not hype, that's clarity. Here are some discussion 00:14:22.100 --> 00:14:26.000 questions for you. And I'll link to the discussion 00:14:26.000 --> 00:14:29.179 in the show notes. What's one task in your day 00:14:29.179 --> 00:14:31.480 that you'd want an agent to handle first? Do 00:14:31.480 --> 00:14:34.019 you trust in AI to plan your calendar? Why or 00:14:34.019 --> 00:14:38.240 why not? Which is harder for you? Planning, remembering 00:14:38.240 --> 00:14:42.159 or fixing mistakes? Share your thoughts and I'd 00:14:42.159 --> 00:14:43.940 love to hear them. That's today's episode, how 00:14:43.940 --> 00:14:46.690 AI agents actually think. If this helped, share 00:14:46.690 --> 00:14:48.570 it with one friend who's drowning in admin work. 00:14:49.009 --> 00:14:50.950 Next time, we'll take it up a notch and talk 00:14:50.950 --> 00:14:53.250 about what happens when one agent isn't enough. 00:14:53.889 --> 00:14:55.950 Multi -agent systems that pass tasks between 00:14:55.950 --> 00:15:00.710 each other. Until then, remember, planning, memory, 00:15:00.870 --> 00:15:04.070 and feedback are the difference between hype 00:15:04.070 --> 00:15:07.710 and something that actually saves you time. There's 00:15:07.710 --> 00:15:10.129 one final discussion point I wanted to cover. 00:15:10.610 --> 00:15:12.879 What's harder for you? Planning, remembering 00:15:12.879 --> 00:15:15.620 or correcting mistakes. Before you go, just wanted 00:15:15.620 --> 00:15:18.659 to let you know that I record with Riverside. 00:15:19.919 --> 00:15:22.279 I'm an affiliate partner of Riverside. They helped 00:15:22.279 --> 00:15:25.159 me record this episode and they have amazing 00:15:25.159 --> 00:15:28.139 audio quality. So that's something I just wanted 00:15:28.139 --> 00:15:30.559 to recommend if you're ever deciding to do your 00:15:30.559 --> 00:15:33.679 own podcast. This is a great platform to record 00:15:33.679 --> 00:15:37.139 because their editing features are amazing. Their 00:15:37.139 --> 00:15:40.980 AI editing is, I think, next level. Helps you 00:15:40.980 --> 00:15:44.139 remove all awkward pauses and gives your sound 00:15:44.139 --> 00:15:48.100 a strong voice. It gives you a strong voice. 00:15:48.840 --> 00:15:52.460 So I would highly recommend Riverside as the 00:15:52.460 --> 00:15:54.320 reason why I chose to be an affiliate partner 00:15:54.320 --> 00:15:56.460 because I actually use the product and I love 00:15:56.460 --> 00:16:01.779 it. So link to join their platform is in the 00:16:01.779 --> 00:16:04.340 show notes. And again, I'm an affiliate partner. 00:16:04.460 --> 00:16:08.120 So if you make any purchase, I may make a small 00:16:08.120 --> 00:16:12.070 commission from it. I also host an rss .com. 00:16:12.570 --> 00:16:15.370 Also, I'm an affiliate partner of RSS because 00:16:15.370 --> 00:16:21.230 they have an amazing distribution channel. They 00:16:21.230 --> 00:16:26.169 are able to publish your episodes from RSS into 00:16:26.169 --> 00:16:29.370 Spotify, Apple Podcasts, Amazon Music, Audible, 00:16:29.629 --> 00:16:35.490 Deezer, Pandora, iHeartRadio, and even let you 00:16:35.490 --> 00:16:38.230 publish your episode with one button into YouTube 00:16:38.230 --> 00:16:42.090 as well. Did I forget to mention that you get, 00:16:42.250 --> 00:16:45.649 you can start making money by hosting on rss 00:16:45.649 --> 00:16:48.450 .com, even if you have just 10 downloads a month. 00:16:49.049 --> 00:16:51.450 So that's all you need to make money, 10 downloads 00:16:51.450 --> 00:16:53.409 a month. It's a new feature they started and 00:16:53.409 --> 00:16:57.009 it's amazing. So again, link to joinrss .com 00:16:57.009 --> 00:16:59.210 will be in the show notes. I'm also using cyda 00:16:59.210 --> 00:17:04.569 .ai to help with my research. So if you are interested 00:17:04.569 --> 00:17:09.160 in any research for your content. or for any 00:17:09.160 --> 00:17:11.700 general you want to learn about anything, want 00:17:11.700 --> 00:17:15.079 to read research papers or any kind of thing 00:17:15.079 --> 00:17:17.960 to improve your knowledge, cider .ai is your 00:17:17.960 --> 00:17:20.759 platform to use. So again, I'm an affiliate partner 00:17:20.759 --> 00:17:23.839 and I'd love if you're able to support the show 00:17:23.839 --> 00:17:27.380 by clicking on my affiliate link and joining 00:17:27.380 --> 00:17:35.359 cider .ai. Hey, it's Mukundan. If this episode 00:17:35.359 --> 00:17:37.369 helped you, Two tiny favors that make a huge 00:17:37.369 --> 00:17:40.990 difference. Rate the show 5 stars. On Spotify, 00:17:41.170 --> 00:17:43.589 you can just open the show page and tap the star 00:17:43.589 --> 00:17:46.130 button. On Apple Podcasts, scroll to the bottom 00:17:46.130 --> 00:17:48.710 of the show page and tap 5 stars. Also, leave 00:17:48.710 --> 00:17:52.369 a one -to -line review on Apple Podcasts. Tell 00:17:52.369 --> 00:17:54.809 me one takeaway. I read every single one and 00:17:54.809 --> 00:17:57.490 it helps more people find the show. That's all 00:17:57.490 --> 00:17:59.210 for this week. Thank you for joining and I will 00:17:59.210 --> 00:18:01.069 see you in the next episode of Data and AI with 00:18:01.069 --> 00:18:02.710 Mukundan, where you learn AI by building.