Help with Telegram + GPT + Plumsail integration (resume matching bot)

Hire a Pro
1

Hi Make Community! :waving_hand:

I’m building a scenario for a Telegram bot that receives a resume (PDF), extracts text via Plumsail, and sends it to GPT (OpenAI) to match it with current job vacancies on our website.

The current structure is:

  1. Telegram receives a document.
  2. Plumsail extracts the text.
  3. HTTP gets live job listings (HTML).
  4. Text Parser extracts job info.
  5. GPT analyzes the CV text and returns relevant job matches.
  6. Telegram replies with the results.

:white_check_mark: What works:
– Resume is received and parsed.
– GPT response is triggered.

:cross_mark: What’s not working:
– The message I get from GPT is just repeated aggregatedText values.
– Seems like GPT is receiving wrong input or the Text Parser doesn’t properly extract job data.
– No clear error handling — users sometimes receive multiple bot replies instead of one.

:bullseye: What I need help with:
– Fixing the GPT input/output formatting.
– Making sure only ONE final Telegram message is sent.
– Optional: better routing & error handling.

:light_bulb: If anyone is open to jumping on a quick call (paid or free) — that would be great. I’m a beginner with Make, but highly motivated and ready to learn. :folded_hands:

Screenshots of the current scenario and bot behavior are available.

Thanks in advance!
– Maksym

1 Like
2

Hello @Maksym_Spirin, welcome to the make Hire a pro section , I would love to work with you on this, you can book a call Here to discuss the project requirements

3

We can help you with this, see: Scalevise – Your Partner in Automations, AI, and Scaling It Right for our previous work

4

Hi, I can help you fix the GPT input/output formatting and make sure your Telegram bot sends only one clean message with accurate job matches. I’ll also review your Text Parser setup and improve error handling and routing as needed.

Happy to jump on a quick call and guide you through the process — I’ve built similar Make.com scenarios with GPT and automation tools.

:backhand_index_pointing_right: Book a call
:link: Upwork Profile

5

Hi Maksym :waving_hand: I’d be happy to jump in and help you fine-tune your Telegram-GPT scenario. It sounds like you’re very close, likely just a few tweaks away from getting it fully functional.

Here’s how we can support:

  • :hammer_and_wrench: Fixing the GPT input by properly formatting job data into structured context.

  • :white_check_mark: Ensuring a single, clean reply is sent back via Telegram (we’ll handle the final step consolidation).

  • :construction: Improving error handling and adding logic to prevent duplicate messages.

  • Bonus: If needed, we can also optimize how job listings are parsed from your HTML (Text Parser can be tricky!).

:paperclip: You can check our work at https://hashlogics.com
:date: Or book a quick call here: https://calendly.com/_hashlogics/30-minutes-meeting

Feel free to DM me, happy to review your screenshots and hop on a call if it helps you learn faster.

6

Hi Maksym,

Great job getting this far — your flow is close! From what you’ve described, it sounds like:

• The job data might not be parsed cleanly from HTML, so GPT is receiving malformed or repetitive input

• The GPT module likely needs better input structuring and context wrapping

• Telegram is sending multiple messages due to either misfiring filters or unbundled routes

I’ve helped build many HR automation systems — in fact, I literally wrote the book on it: The Automated Recruiter: How AI & Automation Are Transforming HR Forever. Happy to walk you through optimizing your GPT prompt structure, bundling logic, and error handling for Telegram.

You can book a time with me here: https://4spotconsulting.com/30

Or check my profile here on Make to grab my email and reach out directly.

Talk soon — and congrats again on diving into Make!

– Jeff Arnold

Founder, 4Spot Consulting

Author, The Automated Recruiter

7

This topic was automatically closed after 30 days. New replies are no longer allowed.