Shadowing Practice: I Wrapped an API Into a Real SaaS (full build / step-by-step) - Learn English Speaking with Video

Choose a practice mode

Each mode opens a full-screen practice room.

338 sentences

If sentences are too short or too long, click Edit to adjust them.

What's going on guys, this is Oliver, formerly from Response AI and a few other software tools

that I've since exited and now running a few new tools and Rosewell.dev, which is where I teach people how to build profitable software,

make money with it, sell it, whatever it may be.

So in this video, I wanted to show you guys how to build a profitable API wrapper.

So in other words, it's really hard to build your own API.

We've talked about this in the past and a lot of different videos.

My last video was about how to build an API.

Some people just make really cool APIs.

They make them for really cheap, and you can just wrap a product around those APIs and sell them for a profit, right?

So we're going to build a keyword monitoring software from scratch, and it's a tool where a user picks a handful of keywords,

and every single day it hunts across social media, drops every new mention straight into their dashboard, and then sort of, you know,

source them in terms of relevance to their website.

So it's almost like Google Alerts, but it actually works.

It's powered by AI.

And the front-end in Cursor and Claude is how it's built.

The back-end is using Superbase.

And the social data that we're scraping from Twitter and LinkedIn, et cetera, is powered by Social Fetch,

which is just an API that pulls live posts from 20-plus platforms with one key basically.

So social fetch is kind of the API that I found to be the most affordable, the fastest as well, because I know I use Appify a lot,

but I kind of prefer this because one, it's just all of its documentation is easier.

You're not relying on, you know, like single developers on Appify to keep up with the,

you know, the Twitter API or the Reddit API, whatever it may be.

And you get every single line of SQL and sort of every piece of code that I write in this video, I'll give to you guys in the description, right?

So where does the data come from?

The one piece you can't vibe code is obviously the actual API that you're tapping into.

You can build the nicest dashboard in the world, but unless you have, you know, posts being fed to it, it's going to be empty.

So scraping social platforms is kind of a nightmare, to be honest.

even I you know running kind of like a lot of

different softwares in the past I've always used APIs for this part

because scraping is kind of a nightmare you get banned you

have to like run loads of expensive proxies it's kind of a pain in the neck right

so tools like social fetch you've obviously got apify you've got

other tools they pull live profiles post search results wherever it may be

and social fetch kind of maintains the scrapers so we don't have to

and we don't have to rely on external APIs or getting banned, right?

And that's what turns this into a bit of an easy project, to be honest.

So what we're actually going to use from the API is the two endpoints, or there's multiple endpoints, but we're going to use X and we're going to use LinkedIn.

So we're going to scrape Twitter for keywords and we're going to scrape X, sorry, LinkedIn for keywords.

And we might even add Reddit at the end.

We might scrape subreddits and then find keywords within those subreddits in posts and comments.

But for now, we're going to keep it kind of simple with just Twitter and LinkedIn.

And the idea, guys, is that you're going to find these LinkedIn posts or these Twitter posts.

And then obviously, if someone mentions something about your product or mentions a keyword that's important to you, then you can reach out to them.

And you can even add another layer to this, which is where you enrich those profiles with some kind of API like Apollo, where you can find their mobile numbers and call them or emails,

you know, and actually email them.

Now, how the product works end to end, guys, the number one is obviously onboarding.

The user signs up and tells us their name, website, offer, then picks up, you know, the keywords that you care about.

And then with a clean pill, input flow, confetti goes off, whatever it may be.

Now, step two is that there's a daily scheduled Cron job, which in other words, is just a job that runs every day.

It's an edge function that takes the user's keywords, searches the last 24 hours on X and LinkedIn through the API with social fetch,

and then writes every fresh mention into their feed with no duplicates, right?

Now, three is the payoff.

The user opens their dashboard every day to a live feed of new mentions about their keywords, platform badges, the tags of the keywords, and filtering.

And it's everyday new results with no effort on their end.

Now, this is a template, guys, for a dozen products, right?

So once you can do this, you can build all of them.

The pattern that you learn today, which is pulling social data on a schedule, storing it in a user's dashboard,

act on what's new, is a skeleton of a whole of a category of tools, right?

So an influencer outreach dashboard that builds a database of creators across TikTok, Instagram, YouTube, etc., pulls their contact emails from their bios,

and then gives you a database right there is a lot of tools out there guys doing this

and making a lot of money a competitor monitor

that watches 20 rival sites and socials daily with the actual endpoints on social fetch

because they can scrape websites as well pings you the instant they ship something new

and that can just be tailored um to you know run with claude

and then claude can just tell you the signals or the changes

or the features updates right the next is something different it'd be like a sentiment trading bot, right?

Which is it watches big trading accounts on LinkedIn or Reddit or Twitter, then reads the mood of every new post or updates or news,

whatever it may be, and then acts on it, right?

It's the same building blocks every single time.

And with the sentiment trading bot, obviously it would just take the actual news or the updates

and then run something like the Binance API to buy crypto

or stock trading API to buy stocks based on news or sentiment.

so in essence really what i'm going to do now is show you how the actual tool works

so this is the landing page right

so know the moment someone's talking about you give echo keywords

your product your name a competitor comma misspelling whatever it may

be every day it searches x on linkedin we've got the

mock-up here um we have got the three steps one page

with every mention build a check in one day right this is just a simple landing page guys

but i'll still give you this code anyway we're going to click start monitoring

and we are going to log in with Oliver Burt, three plus one, and then Echo.

We are going to create an account.

And the first thing that will show us is this prompt.

So what should we call you?

So Ollie, what's your website?

Let's say rosewalt.dev.

It's then going to ask us, what do we sell?

So let's say we sell B2B marketing services for startups.

And then from here, it's going to ask us what keywords, and we've just defaulted this to 10, that we should watch out for.

So let's say if we're selling to, you know, marketing services to startups, let's say cold email, cold calling,

B2B leads, Let's say ZoomInfo and Apollo.io because those are two B2B lead services and sort of data platforms.

Let's say lead gen. Let's say cold outreach and cold outbound.

And then let's try and think of another one.

Let's say B2B SaaS and SaaS, right?

so we've reached the 10 keyword limit and then we can save

and finish there's a bit of confetti

and it says you're all set now this is the feed

where the first sweep of you know keywords comes in at

7 a.m i think mentions will start showing up here once we've searched x

100

and linkedin for the keywords

101

and from here obviously in the top right we can run

102

the sweep now there's also a keywords tab where it shows us what we've added

103

and taken away so we can just take away one and then re-add it and that's all saved to a back end.

104

Obviously, you've got a blank settings and billing, right?

105

So what we're going to do now, we're going to head over to the feed and we're going to click run sweep.

106

And I'm just going to skip ahead to when it's done the sweep.

107

So as you can see, guys, the sweep has actually taken our active keywords inside the edge function that we've set up, which I'll show you how to do in a minute.

108

It's taken all of the queries.

109

It's taken the credits from social fetch and

110

if we just refresh the page again it's finding all of

111

the recent mentions of cold calling cold email etc on twitter

112

and linkedin and it's going to actually populate the feed shortly you can actually now see guys

113

that the feed has been populated with mentions of all of our keywords right

114

so obviously what i would likely add is a um a language filter

115

so that obviously you know things like spanish leads weren't coming through

116

and that kind of thing if we're obviously just in plain English.

117

Obviously, you've got X here and all of the keywords.

118

And then obviously on LinkedIn, you've got the keywords as well.

119

And you can also view the originals of each of these.

120

So a buddy wanted to start cold outbound for his insurance product.

121

We're going to click view original.

122

And obviously, this guy's actually posted about cold outreach in the last three hours, and he helps companies generate consistent pipeline.

123

So you see how that is literally relevant to the keyword itself.

124

And we found a post with different reactions and things like that.

125

So that's how the actual product works and how we could find leads with it.

126

And I'm now going to get into the segment of the video where we actually build this.

127

Just a quick thing, guys, before we actually do build this, I've gone back to add this in.

128

This is the table that is going to be tracking the keywords.

129

You've also got the table that is tracking the monitored posts.

130

And then you've obviously got the profiles, which for these testing purposes is just two of two of mine, basically.

131

Now inside the video and in the description as well, there's going to be the SQL

132

that we're going to be running to make all of these tables so you can just copy me.

133

Okay guys, so the first thing we're going to do is create a super base account.

134

So let's say oliverbert3 plus echo and let's just create a sign up.

135

From there then I presume it's going to ask password is known

136

so annoying man sign up

137

and now what we're going to do is confirm our email confirm the email address

138

and now we have okay there we are

139

so let's just call it echo personal create organization and then And once the organization is created,

140

we just have to use a database password.

141

We'll just say enable data API, and we'll just say it's a European server, which is fine.

142

And we've created a new project.

143

And what we're going to do now is that's going to take a couple of minutes to actually propagate

144

and say it's all healthy and all that, whatever it is.

145

I don't know how fast that happens usually.

146

and now we're going to go back over to cursor right

147

so what I'll do is going to go open folder

148

and I'm going to do a new folder and we're going to call it echo for youtube

149

create and then we're going to open this up and then what we will do is go back over to our PRD.

150

Now the PRD, obviously we've got the product requirements document.

151

We've got the raw voice note as well that I gave to Claude.

152

I'm going to say what this is.

153

I'm going to just pop all this in.

154

Now the next thing we're actually going to do is,

155

or actually should we, I'm going to open the IDE instead,

156

man, because the IDE is like something I'm more familiar with.

157

So Echo YouTube, open this up.

158

There we are.

159

So from here now, we're going to go into Claude Code.

160

And I'm going to paste in the entire PRD.

161

And I'm also going to paste in from Claude the, where are we?

162

Oh my god, this is long, man.

163

We're going to post in the landing page.

164

So copy the landing page.

165

Paste that in.

166

And we're also going to go into Superbase.

167

and we'll run the schema afterwards actually.

168

So where are we?

169

Let's move this out of the way as well. So high effort.

170

I'm going to just paste that in.

171

So now that the entire PRD is being sent into Claude code,

172

I'm just going to take the entire SQL that we created with Claude and run it in Superbase.

173

So that's the wrong one.

174

That's the wrong one.

175

We're going to go into table editor, SQL editor, sorry.

176

Understood, whatever it may be.

177

And we're going to post that in.

178

Now, if we do run, run query, this is obviously 200 lines long.

179

So we're going to go into table editor.

180

And we've got the table for the user's keywords, the monitored posts, the profiles and the sweep runs which is fine

181

we're going to go into authentication i believe it is go down into

182

users or what is it project settings i forget where this is

183

it's not project settings basically what we're doing is turning off

184

the requirement that users have to double confirm their email address.

185

So emails, notifications, confirm sign up.

186

There we are.

187

So sign in, providers, allow new users to sign up and then confirm email.

188

So save changes when we've turned that off.

189

And now obviously I'm probably going to skip ahead

190

because this is going to take ages and Claude is also going to probably ask me to confirm things and deny things.

191

It's already set up our entire project so far

192

and we're just going to wait for it to obviously finish the actual coding session.

193

While that is actually working, we're going to go into the daily sweep index and we're going to copy this.

194

And this is the edge function right so we're going to go into edge functions deploy a new function via editor

195

we're going to call it daily sweep

196

and we are going to paste this entire thing in and deploy the function now that that function is deployed, that is going to be the function

197

that runs every day to check the platforms

198

that we've connected to for the keywords for the user

199

and then send it to their table so that every day when they log in, they see new people across these platforms that have mentioned their keywords.

200

So whilst Claude is actually just running the code in the background, what I'm going to do is go through the full flow with social fetch

201

and how we're using the API and then how we're actually creating something valuable, right?

202

So step one is that user on boards.

203

Step two is that they give keywords they want to track, right?

204

So let's say that this could be, if you were a marketing agency, it could be Google Ads,

205

PPC, Meta Ads, Return on Ad Spend, ROAS, LTV, and CAC, whatever it may be.

206

So all of these keywords are relevant to marketing agencies, right?

207

And obviously the idea is that if someone's mentioning that on Twitter or on LinkedIn, whatever it may be, maybe they're in your market to get your services.

208

So if someone says Google ads, maybe their post or their tweet or their LinkedIn post, whatever it may be, might be saying, I've been trying to run Google ads, but it's unprofitable.

209

So that's the notion is that we can reach out to those people and obviously say, okay, interesting.

210

The same thing applies for, say, if you were a cold email agency, or let's say you were an influencer marketing firm, right?

211

You help brands get more leads with, you know, get more sales with influencers.

212

you could obviously target influencers um return um let's say um sort of cpms instagram leads whatever it may be

213

influencer marketing

214

and again this platform is going to allow us to keep

215

an eye on those people who are mentioning these things

216

and then obviously enrich them and reach out to them

217

and talk to them So the next thing is that every day the edge function runs,

218

finds these keywords online on X and LinkedIn.

219

and we could even add Reddit later,

220

then sends these posts to your inbox inside Echo.

221

So that means that every day you get new people talking about your offer or service online.

222

and there are dozens of tools

223

that do this very profitably right in fact my last tool

224

getseer.com was very profitable it was very very cool people loved it

225

and all it did was obviously just track reddit and twitter

226

and linkedin whatever it may be for people who mentioned these things

227

so i'm effectively cloning an old software that i built that was profitable And then from there, there's unified inbox of all the people.

228

And then later, the super play would be to add an enrichment piece.

229

John mentions cold email on LinkedIn in.

230

We enrich John with something like Apollo or Blitz API.

231

And then that gives us John's email.

232

And then we reach out to John.

233

So, so far, obviously, Echo is only doing this.

234

It finds keywords of people using stuff online or saying stuff online and then gives you their profile.

235

and that's using the social fetch API every single day.

236

What we could do is obviously from here, anyone who comes into your inbox as someone who has mentioned the actual tool

237

or mentioned your services or mentioned they need help, we could then enrich those people with a B2B lead generation API, get their email or their mobile number,

238

and then reach out to them.

239

The same thing applies to, say, a website marketing agency.

240

We're not getting enough leads from our website.

241

That pings over to you, enriches them, and then you can call them or email them and say, I noticed online we're talking about this.

242

So that's the general flow.

243

And that's how social fetch is coming in.

244

And why it's going to be so useful is

245

because it's showing us every day the people who are mentioning that they need our services in so many ways.

246

A super important thing is actually to add the API key

247

from social fetch to the super base edge function secrets so

248

that whenever the edge function is set or the edge function is triggered, obviously we have the API key.

249

So I'm just going to create one on social fetch called YouTube video test.

250

Going to create the key.

251

We've been given the key and we're going to copy it.

252

We're going to head on over to secrets and we are going to create a new secret.

253

Cursor is saying that this is social fetch API key.

254

API key and then we're going to save that in and that is done.

255

Something small I'm doing is actually creating a new Superbase edge function called manual sweep.

256

Now the reason we're doing this is because inside the dashboard

257

the daily sweep is actually scheduled for you know every day at 7 p.m right

258

which we're not going to wait for that just for the video

259

so i'm actually going to create a new edge function

260

that is able to be called by the user at any time

261

and that's just for the sake of the of the testing of social fetch um

262

and to actually you know get us some keywords into the dashboard to show you what we've built

263

but every single day then at um 7 p.m the run is done again on the daily sweep

264

that is the edge function daily sweep running and it will find the new users for that last 24 hours

265

and put them inside your dashboard.

266

Then a couple of things to tighten up guys before the

267

end of the video really is one we're just going to pop into the feed

268

and obviously the daily sweep or the manual sweep has been completed

269

and you can obviously see all of the different posts that have come in.

270

So from here you've obviously got the LinkedIn posts, you've got all of the x posts as well with our keywords etc

271

and that's great and I just wanted to talk quickly about

272

what we could add next to actually improve the value of this.

273

So the two things that I would add is if you go to Blitz API, for example, it's obviously extremely expensive to run these kinds of things,

274

but it's effectively an API library that manages,

275

you know, to get you sort of emails and phone numbers from LinkedIn profiles or from emails, right?

276

So say if you have an email, you can get their mobile number.

277

If you have a LinkedIn profile, you could get their email and their mobile number.

278

If you've just got a name and a domain, for example, like John Evans from johnevans.com, you could get his email and mobile.

279

That kind of thing is obviously really powerful.

280

And that's what we're talking about when we talk about API stacking, right?

281

So what I would likely do is I would add next inside the LinkedIn post or article scraper, which we've added into the feed.

282

obviously roberto vega right so in the original post you've got his actual linkedin right

283

so we've got his linkedin profile you could then obviously feed

284

that to blitz api to get robert vega's email

285

and then send him an email and

286

that would make our um tool echo 10 times more valuable

287

because you're scraping all of the data

288

and then stacking on another data um api

289

which would be to get people's actual mobile and email right

290

that that mobile email to actually do outreach

291

and that's how you could sell this as a software

292

that helps you find people who are talking about your products

293

and then directly message them that's seriously cool right the next thing is

294

that inside the playground for social fetch we could keep going down

295

and you can find reddit right so we're just going to use the playground here

296

so let's go to subreddit and subreddit posts let's go to entrepreneur

297

and let's do the time frame being the last month and the top for the last month

298

and we're going to send this

299

and we have now got the top posts from the last

300

month it's costed us only one credit out of thousands

301

and then obviously we've got all of these posts right the same thing applies to the subreddit search

302

so if we searched into entrepreneur and then search the query cold email and then time frame is of all time.

303

We're obviously going to execute that request as well.

304

And you've got literally inside this post, you've obviously got the most recent posts about cold emails and cold email and burner accounts.

305

Our cold email reply rate went up, et cetera.

306

All of these are obviously posts about the actual keyword.

307

What I would add in is I would use the Reddit search

308

and the Reddit post search and post comment search to improve this by obviously monitoring every single day,

309

Reddit posts and Reddit subreddits so that you could find keywords across those platforms as well.

310

It doesn't even dive into the possibility of, you know, obviously monitoring people who are inside the cold email space,

311

for example, or the the software space on TikTok or on Twitter or on YouTube or on LinkedIn, et cetera.

312

This is just an exciting sort of API or, you know, kind of like product suite to offer.

313

But the one that I would do is I would scrape LinkedIn every day for people who were talking about the product.

314

And then I would use something like Blitz API to enrich their mobiles and emails.

315

And then obviously it goes even further because you could use agent mail then

316

to actually send the actual emails to those people.

317

So imagine a software that you signed up to called Echo, like we talked about and like we built, where you sign up, you give their details, you give your details and the emails,

318

the keywords you want to track.

319

Then it finds all of those people who are mentioning those keywords on the internet, and then it enriches their profiles with emails and mobiles and then sends them a cold email.

320

That is literally what multi-million dollar tools like Artisan are doing and Lavender,

321

or these kinds of tools that are running these kind of outbound motions.

322

Goji Berry, all of these tools are worth millions of dollars and they're doing exactly what we've built today.

323

All you would have to do is tell Claude Co to just add in the Blitz API.

324

You can get a thousand credits.

325

The same thing applies to all of these other different tools that do these sort of mass outbound motions and APIs.

326

The same thing goes for Malt Sets, I believe it is.

327

So Malt Sets is something that you can integrate with to do something like this.

328

Just have a think, guys, about how you could best use this API

329

and feed the data into something else to make a truly sort of valuable product.

330

And obviously, like I said, guys, if you need any questions answered, they'll be in the description.

331

A lot of the documents and things like that, just comment whenever it may be.

332

I'll be there to help.

333

But try out Social Fetch.

334

try out you know building your own tool with this

335

and think about all the opportunities go to social fetch

336

and just feed social fetches landing page to claude

337

and ask it for 10 ideas of how you can make an api wrapper

338

and who you could sell to how you can make it profitable and i'll see you guys in the next video

Download App

Everything you need to speak fluently

AI PronunciationScore every sentence

IPA PracticeMaster every sound

VocabularyBuild your word bank

Vocab GameLearn while playing

Scan QR

Download on theApp Store Get it onGoogle Play

✨ Recommend Video

180 sentences

Your API Can’t Handle Real-World Integrations

Business EnglishUniversity StudentsWorking Adults

190 sentences

This 25-Year-Old Built a 2-Year Moat Nobody Can Bet Against | Corgi, Nico Laqua & Emily Yuan

Business EnglishUniversity StudentsWorking Adults

309 sentences

New Skills! v1.1 brings /wayfinder, /research, /implement, /to-spec, /to-tickets

Business EnglishUniversity StudentsWorking Adults

85 sentences

this will change how you make YouTube videos FOREVER

Business EnglishInterview EnglishUniversity Students

364 sentences

My Vibe Coding Workflow (Step by Step Workflow)

Business EnglishUniversity StudentsWorking Adults

78 sentences

How I deploy serverless containers for free

Daily ConversationUniversity StudentsWorking Adults

Vocabulary and speaking notes for this lesson

This C1 speaking lesson is built on the video “I Wrapped an API Into a Real SaaS”. The speaker keeps coming back to these words: obviously, api, keyword, email, linkedin. This video has 338 sentences and 4558 words to shadow. The speech runs for 27:28. The speaker talks at a natural 166 words per minute, close to everyday conversation. 84% of the words are among the 3,000 most common in English; the rest is worth studying before you start.

Key vocabulary in this video

The 15 most advanced words in the video, with pronunciation and meaning:

Word	Pronunciation	Meaning
keyword noun	/ˈkiː.wɜːd/	Any word used as the key to a code.
fetch verb	/fɛt͡ʃ/	To retrieve; to bear towards; to go and get.
sweep verb	/swiːp/	To clean (a surface) by means of a stroking motion of a broom or brush.
echo noun	/ˈɛk.oʊ/	A reflected sound that is heard again by its initial observer.
scrape verb	/skɹeɪp/	To draw (an object, especially a sharp or angular one), along (something) while exerting pressure.
dashboard noun	/ˈdæʃˌboɹd/	A panel beneath the windshield of a vehicle or aircraft containing indicator gauges, compartments, and sometimes controls and indicator lights.
enrich verb	/ɪnˈɹɪt͡ʃ/	To enhance.
subreddit noun		Any of the subforums, generally intended for discussion of a specific topic, on the Reddit web site.
influencer noun	/ˈɪnflu.ənsə(ɹ)/	A person who or a thing which influences.
blitz noun	/blɪts/	A sudden attack, especially an air raid; usually with reference to the Blitz.
pasta noun	/ˈpɑ.stə/	Dough made from wheat and water and sometimes mixed with egg and formed into various shapes; often sold in dried form and typically boiled for eating.
profitable adjective	/ˈpɹɑ.fɪ.tə.bl̩/	Producing a profit.
outbound adjective	/ˈaʊtbaʊnd/	Leaving or departing; traveling away from; outward bound.
outreach noun	/ˈaʊtɹiːtʃ/	The act of reaching out.
cursor noun	/ˈkɜːsə/	A part of any of several scientific or measuring instruments that moves back and forth to indicate a position.

Phrasal verbs you will hear

Word	Pronunciation	Meaning
sign up verb		To add a name to the list of people who are participating in something.
set up verb	/ˌsɛt ˈʌp/	To make ready for use.
take away verb	/ˌteɪk əˈweɪ/	To remove something and put it in a different place.
try out verb		To test (something) to see how it works or whether it is suitable.
come through verb		To come into a room or other space through a door or passageway.
go back to verb		To stem from; to originate in.
go down verb		To descend; to move from a higher place to a lower one.
go off verb		To explode.

Sentences worth repeating

Short, complete lines from the video that you can reuse in everyday conversation:

It's then going to ask us, what do we sell?
Let's say cold outreach and cold outbound.
Obviously, you've got a blank settings and billing, right?
We're going to click view original.
I don't know how fast that happens usually.

Grammar in this video

The structures the speaker uses most, with the exact words from the video:

Structure	In the video
Present perfect have/has + past participle — a past action that still matters now	We've talked · we've just defaulted · we've reached
Passive voice be + past participle — the focus is on what happens, not who does it	is powered · being fed · be tailored
Relative clauses who / which + clause — extra information about a person or thing	dev, which is · this, which is · today, which is

Pronunciation to watch

The speaker uses 128 contractions and reduced forms, such as we're, we've, I'm. Say them the short way, as you hear them.

The “sh” and “zh” sounds: dashboard /ˈdæʃˌboɹd/, authentication /ɔːˌθɛntɪˈkeɪʃn̩/, refresh /ɹɪˈfɹɛʃ/, notification /ˌnoʊtɪfɪˈkeɪʃn̩/, documentation /ˌdɑkjəmɛnˈteɪʃən/
Long words — get the stress right: influencer /ˈɪnflu.ənsə(ɹ)/, competitor /kəmˈpɛt.ɪ.tɚ/, entrepreneur /ˌɑn.tɹə.pɹəˈnɝ/, authentication /ɔːˌθɛntɪˈkeɪʃn̩/, profitably /ˈpɹɑ.fɪ.tə.bli/

How to practise with this video

Listen to the whole video once without speaking and note the words you do not know.
Start at 0.75× speed, shadow it sentence by sentence, then go back to normal speed once it feels easy.
Record yourself and compare with the original, paying attention to words like keyword, fetch, sweep.

What is the Shadowing Technique?

Shadowing is a science-backed language learning technique originally developed for professional interpreter training and popularized by polyglot Dr. Alexander Arguelles. The method is simple but powerful: you listen to native English audio and immediately repeat it out loud — like a shadow following the speaker with just a 1–2 second delay. Unlike passive listening or grammar drills, shadowing forces your brain and mouth muscles to simultaneously process and reproduce real speech patterns. Research shows it significantly improves pronunciation accuracy, intonation, rhythm, connected speech, listening comprehension, and speaking fluency — making it one of the most effective methods for IELTS Speaking preparation and real-world English communication.

Shadowing technique: read the full step-by-step guide →

Shadowing Practice: I Wrapped an API Into a Real SaaS (full build / step-by-step) - Learn English Speaking with Video

Choose a practice mode

Download App

Your API Can’t Handle Real-World Integrations

This 25-Year-Old Built a 2-Year Moat Nobody Can Bet Against | Corgi, Nico Laqua & Emily Yuan

Claude Code: The Advanced Guide (99% of Devs Skip These Features)

Everyone Is Wrong About AI Coding

The Lazy Method: How I Build $10K/Month Apps

Everything we learned in 29 minutes at Harvard Business School (part 1)

How I'd Get Traffic to a New Website if I Had to Start Over (2026)

A Return to Code

New Skills! v1.1 brings /wayfinder, /research, /implement, /to-spec, /to-tickets

this will change how you make YouTube videos FOREVER

My Vibe Coding Workflow (Step by Step Workflow)

How I deploy serverless containers for free

Vocabulary and speaking notes for this lesson

Key vocabulary in this video

Phrasal verbs you will hear

Sentences worth repeating

Grammar in this video

Pronunciation to watch

How to practise with this video

What is the Shadowing Technique?