CLOUDLESS LABS

Project 08. Technology

Reaching users where they already are, on WhatsApp and Telegram

Integrations with WhatsApp and Telegram, so a system can talk to its users in the messaging apps they already use, and the hard-won lessons about which platform fights you.

What it changed

  • Systems that message users, and take their replies, on WhatsApp and Telegram.
  • Clear advice on which platform suits which job, from building on both.

The situation

Many users will never open your app, but they read their messages. Putting a system on WhatsApp or Telegram lets it notify people, answer questions and take instructions in the place they already spend their day, and with AI agents behind it, that conversation can do real work.

What I built

Integrations with both platforms: webhooks that receive messages, back ends that process them, and replies sent back through each platform's API.

What I learned

The two are not equal:

  • Telegram is the easy one. Its bot API is open and flexible: create a bot, receive updates, reply. Few rules, few surprises.
  • WhatsApp is the hard one. Templates need approval, the rules on what you may send and when are strict, and a number can be banned quickly if the platform doesn't like how it's used. It's worth it when your users are on WhatsApp, but plan for the compliance work and protect the number.

Why it matters to you

If you're choosing a channel for a bot or an agent, the platform's rules matter more than its API. I'd start on Telegram when the audience allows it, and go to WhatsApp with eyes open when they don't.

Related projects

  1. 11 Crypto market analysis and an automated trading bot Personal research project · Finance Quantitative research on crypto markets and an automated trading bot built on it: market data, signals, back-testing and execution, around the clock.
  2. 01 Off a fragile VPS and onto AWS, with almost no downtime An Australian agriculture-technology company with a large IoT sensor network · Agriculture technology A growing agtech platform ran on a single server that could neither scale nor be secured. I moved all of it to AWS, live, with almost no downtime, then rebuilt it to scale.
  3. 02 Cutting an AWS bill from US$240k a year to under US$40k An Australian agriculture-technology company with a large IoT sensor network · Agriculture technology · In progress An agtech platform was paying about US$240,000 a year for AWS. The programme under way takes it under US$40,000: roughly US$200,000 a year back in the business, for the same workload.
  4. 03 Real-time ingestion for a national farm-sensor network An Australian agriculture-technology company with a large IoT sensor network · Agriculture technology Thousands of field devices reporting around the clock, processed as the data arrives instead of hours later, on infrastructure that scales with the network.

Got a project like these?

Tell us where the business needs to go. You'll talk to the engineer who will design and build it, not a salesperson, and get an honest answer on whether we're the right fit.

Contact Us

For general inquiries, please contact us at:

nic@cloudlesslabs.com

For e-learning capabilities, please contact:

lin@shirleenwong.com