YouTube2Text

Veja Como eu Criei Um Assistente de Consultas Processuais Com IA no N8n — Transcript

by Vitor Pasqualotto · 3,588 words · 510 segments · language en · Watch on YouTube

Full transcript

  1. 0:00In today's video, I'm going to show you
  2. 0:03how we can build an AI agent that looks
  3. 0:05up legal cases based on the case number
  4. 0:08by accessing a public API and helps you
  5. 0:10analyze them by sending you a summary
  6. 0:12of the case, okay? And chatting with
  7. 0:16you about the matters regarding that
  8. 0:18case. So, if this is a topic that
  9. 0:21interests you, stay until the end of
  10. 0:22the video, and I guarantee you'll like
  11. 0:24it. Today will be a slightly shorter
  12. 0:27video, but it will be a very useful one
  13. 0:29. For those in the legal field who have
  14. 0:31to deal with cases and constantly look
  15. 0:33up this kind of thing, you'll like the
  16. 0:35video and find it interesting, because
  17. 0:38I think it's a flow that's very simple
  18. 0:40but can also help you a lot in your
  19. 0:41day-to-day, okay? So, without further
  20. 0:44ado, let's get to what matters, let's
  21. 0:46go to the computer screen. So, alright,
  22. 0:48here on my screen you can see the agent
  23. 0:50, it's already built, okay? A very
  24. 0:53small agent, where it's connected to
  25. 0:56just one tool, but it's interesting
  26. 0:59because even though it's simple, it can
  27. 1:02be very useful, okay? Like I mentioned,
  28. 1:06for those who work with this daily and
  29. 1:08have to keep looking up cases all the
  30. 1:11time. And this could be a very
  31. 1:13interesting flow for you, okay? That
  32. 1:15will help you save a good amount of
  33. 1:18time, okay? All right? The flow of this
  34. 1:21agent is very simple. Basically, we'll
  35. 1:24start the flow with the chat trigger
  36. 1:26from n8n itself, okay? We'll connect
  37. 1:30our agent to an AI model, right? In
  38. 1:33this case, I'm using GPT itself, I'm
  39. 1:36using the 4o model, a memory, and a
  40. 1:39tool, an HTTP request tool, because
  41. 1:42we're going to query a public API, okay
  42. 1:45? And to retrieve this case information
  43. 1:49. Got it? So, before we build this
  44. 1:52agent step-by-step, let's take a look
  45. 1:54at this API so you can understand more
  46. 1:57or less how it works, okay? And we'll
  47. 2:00also perform some basic tests using
  48. 2:04Postman, okay? To test these requests
  49. 2:08to this API. Got it? So here we can
  50. 2:13search for DataJud, okay? It's possible
  51. 2:16that some of you already know this API.
  52. 2:20Alright? Here we can take a look, right
  53. 2:23, at the public API, have a read here
  54. 2:26to understand how it works. The DataJud
  55. 2:29public API is a powerful tool that
  56. 2:31offers public access to metadata from
  57. 2:33legal cases throughout Brazil. The data
  58. 2:36made available through the API comes
  59. 2:38from the national judiciary database,
  60. 2:40Datajud. They strictly adhere to the
  61. 2:43criteria stipulated by ordinance
  62. 2:45Tananã Tananã, while simultaneously
  63. 2:47ensuring the protection of confidential
  64. 2:49cases and the information of the
  65. 2:51parties involved. So, basically, it
  66. 2:54will provide us with case data
  67. 2:55throughout the entire Brazilian
  68. 2:57territory, right? You were able to
  69. 2:59understand this part, right? I hope so.
  70. 3:03We can access this link here, where it
  71. 3:07will take us directly to the public API
  72. 3:11documentation, okay? So, we can close
  73. 3:15that other tab, and here it says I
  74. 3:17think more or less the same thing we
  75. 3:19already read, okay? Through this API,
  76. 3:23developers have access to information
  77. 3:25on case headers and movements related
  78. 3:27to judicial proceedings in all
  79. 3:29instances of the Brazilian judiciary.
  80. 3:32All good? Here we have some terms of
  81. 3:35use, which I recommend you read, but if
  82. 3:37you just want to run some small tests,
  83. 3:39that's fine, I don't think there's any
  84. 3:42problem. And access. Here we will
  85. 3:47understand how we are going to hit the
  86. 3:48endpoints of this API, okay? How will
  87. 3:51we be able to access it to successfully
  88. 3:53return the data we want? The Datajud
  89. 3:57public API, the tool that provides
  90. 3:59public access. The same thing we
  91. 4:01already read. API key here.
  92. 4:04Authentication for the Datajud public
  93. 4:06API is performed using a public key,
  94. 4:08generated and provided by the DPJ CNJ.
  95. 4:12The current key is always accessible in
  96. 4:14this section of the Wiki. So, basically
  97. 4:16, we will use this key here. It is a
  98. 4:19public key that anyone can come here
  99. 4:22and grab, okay? So, basically, how are
  100. 4:25we going to do it, to use this key,
  101. 4:28okay? We will pass an authorization key
  102. 4:33in the headers of our request, and the
  103. 4:35value will be "API Key" followed by the
  104. 4:38key, okay? So, basically, this is how
  105. 4:43we will hit our requests. To start, we
  106. 4:47can already configure these headers in
  107. 4:49the request we are going to make. I
  108. 4:52have Postman open here already, and I
  109. 4:55have two endpoints separated here. No,
  110. 4:58two endpoints, it's the same place we
  111. 4:59are going to hit. We are just going to
  112. 5:00change the method. In one we will send
  113. 5:02the body and in the other we won't,
  114. 5:03okay? This first one here, what is this
  115. 5:07GET for? We are just going to hit the
  116. 5:10API, okay? In a specific endpoint, and
  117. 5:13it will return some cases for us. So
  118. 5:15let's run this test here, okay? But
  119. 5:18first we need to know the URL, right? I
  120. 5:22forgot to show you that. We can go here
  121. 5:26to endpoints, and what we'll use is
  122. 5:28superior courts and superior court of
  123. 5:30justice, okay? STJ. But you're not
  124. 5:35going to click here, okay? We're going
  125. 5:38to copy this link, all right? We can go
  126. 5:42back here to Postman and we can paste
  127. 5:45it here in the URL, okay? And here in
  128. 5:48headers, we'll pass an authorization
  129. 5:51and we can go back here to access and
  130. 5:54copy all of this. Got it? We pass it
  131. 5:59here in value and we can send this
  132. 6:01request now. You saw that the status
  133. 6:04code 200 returned everything correctly.
  134. 6:06It will return a ton of lines here,
  135. 6:08okay? About how many? 1,000 lines. I
  136. 6:12thought more lines would come, but
  137. 6:14that's fine. If we search by case
  138. 6:16number, we'll be able to see that we
  139. 6:19can find 10 cases here, okay? 10
  140. 6:22different case numbers, and each case
  141. 6:25has some specific information. Uh, for
  142. 6:28example, we have the class, we have
  143. 6:31this class name which, if I'm not
  144. 6:34mistaken, is the description of the
  145. 6:36case class, uh, format, the court, the
  146. 6:39last update date, the movements of this
  147. 6:42case, okay? So here we can check all of
  148. 6:45that. Any judge, all right? Here it's
  149. 6:53basically that, there are movements and
  150. 6:55also the subjects of the case, okay? So
  151. 6:57we will have access to all this data.
  152. 7:00Why am I hitting this endpoint?
  153. 7:01Basically, I just hit it because I want
  154. 7:04to get a specific case, a number of a
  155. 7:06specific case so we can test, okay? So
  156. 7:08I'll take this case right here, got it?
  157. 7:11And what we're going to do is hit this
  158. 7:13endpoint by sending a body, okay? And
  159. 7:15in this body, we'll pass the case
  160. 7:17number. By passing the case number, it
  161. 7:19will return to us only the data for
  162. 7:21that specific case. And that is what we
  163. 7:23want. Our chatbot works like this.
  164. 7:27Basically, we'll have access to a case
  165. 7:29number and we want to know its
  166. 7:30information. We want a detailed summary
  167. 7:34of this case. So we will send the case
  168. 7:37number to our agent so it can query the
  169. 7:39API. Then it will return the response.
  170. 7:43The response will basically have a
  171. 7:44summary, a small analysis of this case,
  172. 7:46okay? Got it? So we can copy this URL
  173. 7:50here, okay? We can go over to the
  174. 7:52headers here, it will be the same thing
  175. 7:55. Authorization. And here we can copy
  176. 7:59this as well, okay? All good? Now here
  177. 8:03in body we will select raw, okay? This
  178. 8:05is a JSON. And we can go back here. If
  179. 8:10we take a look at these examples. we
  180. 8:12can see here, okay? Uh, it even
  181. 8:15mentions here, it recommends Postman.
  182. 8:18Ta-da, we're going to hit it with the
  183. 8:20post method, right? We are sending a
  184. 8:23body. And basically here we will use
  185. 8:27this example body, okay? This body that
  186. 8:30they left as an example here. So, we
  187. 8:32can copy this and we can paste it right
  188. 8:34here, okay? The only difference is that
  189. 8:37we will have to replace this case
  190. 8:39number that is already in the example
  191. 8:41and put the one we just got, okay? So
  192. 8:44I'm going to grab that. If you have any
  193. 8:47number, if you have access to any case
  194. 8:49number there, feel free to test it, put
  195. 8:51it in and send the request, okay? Got
  196. 8:54it? It has to be in this format. So
  197. 8:57here we can hit send and it's OK,
  198. 8:59everything is just right. Here you will
  199. 9:02see that it returns many fewer lines,
  200. 9:04right? about a hundred and something.
  201. 9:07So it's just the information for this
  202. 9:09specific case, alright? So, with the
  203. 9:12API tests done via Postman, we can go
  204. 9:14to n8n and start to understand how this
  205. 9:17flow works, and we'll also build it
  206. 9:19from scratch so you understand it
  207. 9:21step-by-step, okay? So I'll minimize my
  208. 9:25Postman here, alright? Here too, if you
  209. 9:28want, you can take a look at the other
  210. 9:30examples, okay? There are other ways
  211. 9:32for you to query the API, there's
  212. 9:34something else here, a data glossary,
  213. 9:35okay? So it has the data that will
  214. 9:38return in the response of the request
  215. 9:41we sent, okay? So, court, case number,
  216. 9:44filing date, degree, secrecy level, uh,
  217. 9:47class, name, oh, the class name I had
  218. 9:50mentioned to you, description of the
  219. 9:52procedural class. So, it explains each
  220. 9:56element here, okay, of the response you
  221. 9:59might receive. final. There are a few
  222. 10:03more little things here. Feel free to
  223. 10:05read through all this documentation
  224. 10:07here so you can understand how it works
  225. 10:08, okay? Eh, so, with all that said,
  226. 10:10with that explained, let's go to n8n,
  227. 10:12okay? Let's go back there, because
  228. 10:14that's where the magic really happens.
  229. 10:16So, very simple. We have a chat trigger
  230. 10:19, right? And here we have our agent
  231. 10:23itself, where we will pass a small
  232. 10:26prompt, okay? Explaining to it what it
  233. 10:29has to do. We pass the chat input, it
  234. 10:31will consult the tool, perform the
  235. 10:34query, and return the answer to us,
  236. 10:36okay? Basically, if we were to build
  237. 10:40this from scratch, we would pick a
  238. 10:43trigger. Where is it? We can't. Let me
  239. 10:50do the following, let me delete this.
  240. 10:55And here we will select a chat on chat
  241. 10:57message, okay? It's no secret, this
  242. 11:00isn't news to anyone anymore, right? So
  243. 11:04you can select it, we can connect it to
  244. 11:06our agent, okay? But in this case, I
  245. 11:08will pull it down here. Advanced AI. AI
  246. 11:13agent. Ah, where is it? It pulled it up
  247. 11:18there, but that's no problem. Here I
  248. 11:21will copy the prompt, okay? Just to
  249. 11:24avoid wasting time, but I will read it
  250. 11:26with you so you understand. I will pass
  251. 11:29a system message here just explaining
  252. 11:31that it is a very useful and
  253. 11:32intelligent legal agent, okay? Basic
  254. 11:35stuff here. The main prompt we can
  255. 11:39actually define here as defined below,
  256. 11:44okay? Expression. We can open this
  257. 11:48little screen here and copy the prompt.
  258. 11:50You should talk to the user about cases
  259. 11:52; when the user asks you to check a
  260. 11:54case, you must identify the provided
  261. 11:55number and consult the Tool, which we
  262. 11:57have named for this Tool. retrieve
  263. 11:59information about the case in question.
  264. 12:02If the user asks for a check but didn't
  265. 12:04provide the number, ask them to do so
  266. 12:07so you can perform the search correctly
  267. 12:09. When you find the case, perform an
  268. 12:12analysis of the content the API returns
  269. 12:14and synthesize all the content of the
  270. 12:16case in question. Use a friendly, nice
  271. 12:19tone, with a touch of humor, okay? You
  272. 12:21can define whatever you want here. And
  273. 12:23next, for the user input, you pass a
  274. 12:25JSON chat input here, got it? So that
  275. 12:28whatever you type in the previous
  276. 12:30trigger module, it will send over here.
  277. 12:34All good? There's no secret here,
  278. 12:36there's nothing else to do, okay?
  279. 12:38Basically, it's just the prompt. From
  280. 12:41here we can pull our model, okay? And
  281. 12:44here we can pull the OpenAI chat model
  282. 12:47or whatever you want to use, okay? You
  283. 12:50can set up your account with your
  284. 12:54correct credentials. Here I will use
  285. 13:00the 4o. I'm not sure if this is the new
  286. 13:07version of N8n, but that's fine. Here
  287. 13:10it gives a huge list and we can't
  288. 13:12search, it closes, but that's fine.
  289. 13:16Configure the model you want to use
  290. 13:18here, and the memory, okay? Window
  291. 13:21buffer memory is quite traditional,
  292. 13:24simple to configure. And our tool,
  293. 13:28finally, right, so you can see that
  294. 13:31it's a ridiculously easy agent to
  295. 13:34configure. It's an HTTP request tool.
  296. 13:38So here we can provide a description. I
  297. 13:41have already written the description
  298. 13:42for this tool. Copy this here, okay?
  299. 13:47Great. Uh, yeah. Great. I'm just going
  300. 13:55to quickly copy this URL, okay? So
  301. 13:58let's configure it here. The method we
  302. 14:02are going to use will be a POST. Got it
  303. 14:04? For the URL, we provide the correct
  304. 14:08URL we used a moment ago in Postman.
  305. 14:13Here we can select sender, okay? You
  306. 14:16can activate this here. Uh, it can be
  307. 14:19in these fields below. We will type
  308. 14:23authorization, okay? Here we can select
  309. 14:28using field below. And here we provide
  310. 14:31the value, okay? Let me go back here to
  311. 14:34Postman in headers. Let's copy this
  312. 14:37here. Go back here. Provide the value.
  313. 14:40Remembering to copy this API that is in
  314. 14:43front of the key, okay? Send. Here I
  315. 14:48will, we will use a JSON, okay? So, use
  316. 14:51JSON below. And here we will use this
  317. 14:55same body that we used here, okay? With
  318. 15:01a single difference here in the process
  319. 15:04number, what are we going to use? If
  320. 15:08you look here at the top, if I'm not
  321. 15:11mistaken, it's placeholder. For any
  322. 15:15part of the request to be filled by the
  323. 15:17model. So, we can use this syntax here
  324. 15:21inside braces where we will put a like
  325. 15:26it's a variable name, okay? Here it's
  326. 15:30placeholder. So we will put here the
  327. 15:32process number, okay? Process number.
  328. 15:39I'll even copy this so there are no
  329. 15:41mistakes. And down here we will put in
  330. 15:45placeholder definitions, right? We will
  331. 15:48set a placeholder. We can put its name,
  332. 15:50okay? Which is the same one we put here
  333. 15:52between this pair of braces. And we
  334. 15:54will provide a description for this
  335. 15:56placeholder. I will again copy this
  336. 15:58description that I've already written
  337. 16:01here. So you provide a little
  338. 16:06description for this placeholder so we
  339. 16:08understand what it needs to fill in
  340. 16:10there, okay? So it's the process number
  341. 16:13you wish to consult, which you can set
  342. 16:15as a string, okay? And that's it. Now
  343. 16:19on to the testing phase. Very simple,
  344. 16:21right? So let's open a chat here and
  345. 16:24grab a case with a different number.
  346. 16:28Let's pick, I don't know, seven. Uh,
  347. 16:35first let's say hello, right, to test
  348. 16:38its response. Oh, hello. How can I help
  349. 16:44you today? If you need any case
  350. 16:46consultation or have questions about a
  351. 16:48legal matter, I'm here to help. So, I'd
  352. 16:51like you to summarize the following
  353. 16:57case. We passed it a different case.
  354. 17:02Let's see if it works. Apparently it
  355. 17:05worked. The API returned a response
  356. 17:09here. Done. Sure. Let's take a look.
  357. 17:15This is an ordinary appeal in habeas
  358. 17:17corpus registered at the STJ. It is
  359. 17:21completely electronic because we are,
  360. 17:24right? The action was filed on November
  361. 17:2711, 2024. Responsible body, office of
  362. 17:30Minister Sebastião Reis Júnior.
  363. 17:33Alright. Main case movements.
  364. 17:36Distribution occurred, it was done by
  365. 17:38lottery. Conclusion for decision. Right
  366. 17:41after that, the case was set for
  367. 17:42decision on the same day. Issuance of
  368. 17:44documents and petitions. There were
  369. 17:48several, including merit opinions from
  370. 17:50the Public Prosecutor's Office and
  371. 17:52other important petitions. Alright? So
  372. 17:55it gave the main movements of this case
  373. 17:57here, right? What was going on?
  374. 18:01Subjects involved, drug trafficking and
  375. 18:04related conduct, and provisional
  376. 18:06release. So here was a very simple
  377. 18:08little response, right? We are going to
  378. 18:10check other cases here to test. And
  379. 18:14there is a judicial journey with many
  380. 18:16emotions and documents flying
  381. 18:17everywhere, virtually, of course. If
  382. 18:19you need any more details or have other
  383. 18:21questions, I am here. So, as you could
  384. 18:24see, it responded very well, right?
  385. 18:25Basically, it synthesized this case
  386. 18:28that the API returned to it. And let's
  387. 18:31test two more cases here, okay? Just so
  388. 18:33we don't have any doubts or anything
  389. 18:38left. Take this second one here. Copy
  390. 18:41this number now. This one. Let's see.
  391. 18:52It hit the tool. We can verify the logs
  392. 18:55. The API returned the response. I
  393. 18:59think this one might be a little bigger
  394. 19:00than the other. or not? Uh,
  395. 19:04distribution, it was distributed on
  396. 19:07December 19, 2024. On the same day, it
  397. 19:11was forwarded for decision. Alright.
  398. 19:15Alright, done. This case is supervised
  399. 19:21by the presidency of the STJ and its
  400. 19:24last update was quite recent, March 10,
  401. 19:262025. It seems our dear 13th salary is
  402. 19:30causing quite a stir in the courts. If
  403. 19:32you need more information and have any
  404. 19:33questions, I will be happy to help.
  405. 19:36Alright? Now, finally, let's consult
  406. 19:38one more case, right, so we can see the
  407. 19:40response it will return. Let's pick
  408. 19:42five. Let's wait. Pretty fast, right?
  409. 19:59His, his answer. Great. And this is an
  410. 20:06interlocutory appeal of a special
  411. 20:07appeal. It is entirely electronic,
  412. 20:09great for saving some trees. Uh, main
  413. 20:12updates. Distribution, September 10,
  414. 20:15'24, remittance for redistribution. Ah,
  415. 20:19receipts were recorded on the 3rd,
  416. 20:21September 23, 2024. Conclusion for
  417. 20:25decision. Matters involved. Violation
  418. 20:27of administrative principles. Burden of
  419. 20:30proof. Who has the job of proving what?
  420. 20:33Uh, I don't know much about law, okay?
  421. 20:35I'm just reading here. So, for those
  422. 20:37who understand, cool? Ah, obligation to
  423. 20:40do, or not to do. Someone being ordered
  424. 20:42to take actions or inaction. Damage to
  425. 20:46the treasury and illicit enrichment.
  426. 20:49This touches the government's wallet
  427. 20:51and perhaps others 'too. Great.
  428. 20:54Minister Afrânio Vilela. This case is
  429. 20:59full of discussions about
  430. 21:01administrative principles,
  431. 21:03responsibility, and even some complex
  432. 21:06financial issues. Cool, right? So, as
  433. 21:10you could see, it, it responds very
  434. 21:13well, okay? So, if you need to consult
  435. 21:16cases all the time or you, I don't know
  436. 21:18, saw a case but don't remember exactly
  437. 21:20what it's about, if you put it here,
  438. 21:22you'll get a quick summary of this case
  439. 21:24. You can ask it to summarize even more
  440. 21:27or provide even more details. So, all
  441. 21:30of this you can specify in your prompt,
  442. 21:32okay? Any specific little detail or
  443. 21:34some way of responding or let's see
  444. 21:37some response structure that you want
  445. 21:40the agent to use, you can send it to us
  446. 21:43, okay, in the prompt itself and it
  447. 21:45will know how to answer you in the best
  448. 21:48possible way for you, okay? So, if you
  449. 21:51are in the field, if you are a lawyer
  450. 21:53or whatever it is in the legal area, uh
  451. 21:55, you understand more about the subject
  452. 21:57, you can specify all of that in your
  453. 21:59prompt, okay? The way you want it to
  454. 22:01respond. You can mix this with perhaps
  455. 22:04a few more tools. If you want to use
  456. 22:07some knowledge base for your agent to
  457. 22:09use, you can also configure that type
  458. 22:11of flow. We can configure this on
  459. 22:16Telegram and keep the scenario active
  460. 22:18so you have an agent always available
  461. 22:21to you on Telegram, where you can
  462. 22:23consult cases at any time of the day,
  463. 22:26from your own cell phone, okay? So
  464. 22:29there are many use cases, many
  465. 22:31scenarios, okay? That you can use this
  466. 22:33in the best way. That's what's
  467. 22:35important, right? After all, these
  468. 22:37agents, these flows serve to improve
  469. 22:39our work, right? Ah, so alright. Yeah,
  470. 22:43that was what I wanted to show in this
  471. 22:44video. It's an extremely simple
  472. 22:46workflow, okay? Anyone can come in and
  473. 22:49set this up in a little over 5 minutes,
  474. 22:51I don't know. So that's it, take a look
  475. 22:54at the documentation, ah, create this
  476. 22:57flow yourself, put it into practice,
  477. 22:59right? Especially if you're in the
  478. 23:01field, put it into practice. Maybe it
  479. 23:03will help you, maybe not. Eh, it
  480. 23:05depends on what you do in your work,
  481. 23:08right? If you work directly with this,
  482. 23:10it might help you. Otherwise, there are
  483. 23:12several other workflows we can build
  484. 23:14focused on the legal field, right? And
  485. 23:18if you guys really like this type of
  486. 23:20content or if there are any lawyers out
  487. 23:22there or people in the legal field who
  488. 23:24like my videos, leave a comment. I can
  489. 23:27bring other workflows, okay? Related to
  490. 23:30this area. Eh, and there is an infinity
  491. 23:33of possibilities, okay, that we can, eh
  492. 23:36, build focused on law or any other
  493. 23:39area, okay? Leave suggestions down
  494. 23:42below as well, okay? Your opinion is
  495. 23:44always very welcome, your comment, your
  496. 23:45feedback, okay? It helps a lot, it
  497. 23:47gives new ideas. And that's it, this
  498. 23:50video is coming to an end. Leave your
  499. 23:52like. If you liked the video, subscribe
  500. 23:54to the channel, if you're not
  501. 23:55subscribed, turn on notifications.
  502. 23:57Every week there's a new video, okay?
  503. 23:58I'm looking to increase the frequency
  504. 24:01of videos to, you know, bring more
  505. 24:02topics to you, right? The channel is a
  506. 24:05bit slow, but there are some projects
  507. 24:08underway and good things are coming in
  508. 24:10the future, okay? So, that was what I
  509. 24:13had for today's video. Deal? So, I'll
  510. 24:16see you in the next one.

About this transcript

This page contains the full transcript of Veja Como eu Criei Um Assistente de Consultas Processuais Com IA no N8n by Vitor Pasqualotto, generated from the public captions YouTube serves with the video. The transcript has 3,588 words across 510 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.