AI bot for call assessment (Ringostat + Kommo CRM) + refinement of the video meeting bot (Sembly)

Good day! I present the Nexus Core team.
For implementation, we use the following stack and architecture:

Backend: Node.js / Nest.js for integration with Ringostat API and Kommo CRM.AI
Processing: OpenAI Whisper for transcription and GPT-4 for evaluating dialogues based on specified criteria.
Data Flow: Automated transfer of results (scores, recommendations, agreements) to Google Sheets or PostgreSQL database.
Logic: Separation of scenarios for video meetings through custom prompts and processing conditions for the 1st and 2nd meetings.

Clarifying questions:
… Does Ringostat provide recordings in separate channels (manager/client) to improve analysis accuracy?
How is the existing bot for video meetings currently technically implemented (language, API)?
Are automatic notifications needed for the manager in Telegram in case of a low score?

Cases under NDA, but we are ready to discuss in detail the implemented projects of a similar class in private messages. Write in private messages — we will clarify the details and provide a fixed cost with a breakdown by stages and agreed deadlines.

Show original

Translate

Good day! I specialize in automating business processes through AI and API integrations (Ringostat, Kommo, Sembly). I offer a solution: a service on Python/n8n that converts call recordings into text, analyzes them using LLM, and structures the report in CRM and tables. For video meetings, I separate the evaluation logic through individual AI prompts. I am ready to start working soon.

Show original

Translate

I specialize in creating AI systems for quality control in sales departments and I am in the top 5 specialists in the category of AI and machine learning. Your task is very clear to me, as I have already implemented similar solutions specifically with the Ringostat + Kommo CRM combination.

My vision for the architecture of the solution:
Call evaluation bot (Ringostat + Kommo):

Data collection: I will set up automatic retrieval of call recordings from Ringostat via Webhooks immediately after the conversation ends.

AI analysis: Using Whisper models (transcription) + GPT-4o/Claude to evaluate according to your checklist. The system will not just look for words, but will assess meanings, intonations, and adherence to sales stages.

… Reporting: Data will instantly go to Google Sheets (employee, scores, strengths/weaknesses, recommendations) and will be duplicated in the deal card in Kommo CRM for the convenience of the sales manager.

Refinement of the video meeting bot (Sembly):

Logic distribution: I will set up triggers that will determine the type of meeting (1st or 2nd) and apply the appropriate checklist.

Scaling: I will update the scenario architecture (based on Make/n8n) so that adding new checklists does not disrupt the existing logic.

I am ready to discuss the details and show examples of reports in private messages. When would it be convenient for you to talk?

Show original

Translate

Good day. I am ready to complete this project as I have extensive experience in application development.

Show original

Translate

Good day.

I have reviewed the task description. Similar call and meeting evaluation systems are usually built through a pipeline: obtaining the recording → transcription → AI analysis according to a checklist → structured report → transfer to CRM/spreadsheet.

Here is how I see the architecture of the solution:

1. Obtaining recordings
Integration with the Ringostat API to obtain call recordings and metadata (date, manager, client number, duration).

… 2. Transcription of the conversation
Converting audio to text (Whisper / another STT model).

3. AI analysis according to the checklist
LLM analyzes the transcript according to specified criteria and generates:
— a score for each item
— strengths
— weaknesses
— agreements with the client
— recommendations for the manager.

4. Formation of a structured result
The result is formed in the form of JSON and transferred:
— to a spreadsheet (Google Sheets / database)
— to Kommo CRM if necessary.

5. Refinement of the video meeting bot
It is necessary to separate the evaluation logic:

— scenario of the 1st meeting
— scenario of the 2nd meeting

and implement separate checklists for each scenario with correct result transfer.

Estimated cost:

Development of the call evaluation bot
$400–600

Refinement of the video meeting bot
$150–300

Implementation timeframe:
7–12 days depending on API availability and the current implementation of the existing bot.

I am also ready to initially conduct a short technical audit of the current system to more accurately assess the scope of refinements.

I would be happy to discuss the details.

Show original

Translate

Hello! I have experience working with the Ringostat and Kommo APIs, so I understand the logic of data transfer in this connection. I propose to implement the project in Python (FastAPI/Pyrogram), which will give you full control over processing without monthly payments for low-code platforms. Here’s how I envision the implementation:

Call Evaluation (Ringostat + Kommo):

- Processing: I will use Webhooks from Ringostat. As soon as the call is completed, the script will retrieve the audio, transcribe it using Whisper (or Deepgram for faster processing), and send the text to GPT-4o.
- Checklist: I will set up the system so that the AI produces strictly structured JSON. This ensures that scores, strengths, and recommendations will always go into the correct columns of the table and fields of the CRM without errors.
- Flexibility: You will be able to edit the criteria in Google Sheets without accessing the code - the bot will pull updated requirements from there.

Refinement of Sembly (video meetings):
…
- Logic: I will separate the scenarios based on the input data levels. The system will recognize the type of meeting (1st or 2nd) by tags or the meeting name in Sembly and connect the appropriate prompt and checklist.

- Result: Data from both types of meetings will be correctly separated in the database and final reporting.

Why Python instead of n8n/Make: for processing long recordings and complex checklists, code is more reliable. You will not have limits on the number of operations or data volume, and the system will work faster due to asynchronicity.

I am ready to discuss the details, feel free to write.

Show original

Translate

Good day!
I am ready to take on: a call evaluation bot (Ringostat + Kommo) + improvements to the video meeting bot (Sembly).
I will do everything on n8n - quickly, reliably, without unnecessary code.
Cost: 18,000 UAH (with one full round of revisions after tests).
Timeline: 12-14 working days after access and checklists.
Architecture briefly:
Separate workflows in n8n: one for calls, two for meetings (or one with IF branches).
Calls: webhook/polling Ringostat → transcript (their own or Whisper) → LLM (Groq/Claude/OpenAI) according to the checklist → JSON → Google Sheets / Kommo.
Video: webhook Sembly → transcript → IF (1st or 2nd meeting) → different prompts/checklists → the same output in a table.
… Advantages of n8n: easy to change logic, retry, logs, cheap hosting.

I am ready to sketch a scheme one day after the checklists.
I am waiting for a response!

Show original

Translate

Good day. I have experience in developing AI bots and integrations with CRM and API services. I can implement a bot for analyzing Ringostat calls, generating evaluations based on a checklist, and transferring the results to a table/CRM, as well as refine the logic for evaluating Sembly video meetings.

I will create a stable architecture (record processing → AI analysis → report generation → integration with CRM).
I am ready to start quickly and discuss the details.

Show original

Translate

Good day.

I have experience in developing AI solutions for analyzing calls and video meetings with CRM integration and transferring structured reports into tables/databases. I have worked with transcription, checklist evaluation, and the formation of scoring models.

I suggest moving to private messages for clarification of the technical specifications and final cost calculation.

Show original

Translate

Good day!

I have experience with AI bots, APIs (Ringostat, Kommo, Sembly), and Google Sheets. I am ready to develop a call evaluation bot and refine the video meeting bot.
Architecture: Python service, integration via API, AI/NLP analysis, result in Google Sheets.

Message me privately, and we will discuss the details, estimated cost, and implementation timelines.

Show original

Translate

Hello!
I am engaged in the automation of business processes and the development of AI bots on the n8n platform. I have implemented an AI call analysis system with CRM: the bot receives recordings, makes transcriptions, analyzes the conversation according to a checklist, generates an evaluation, recommendations, and sends the results to CRM and Google Sheets.
1. Call evaluation bot (Ringostat + Kommo)
It will automatically receive recordings, analyze conversations based on your criteria, and generate a structured report: employee, date, client, duration, strengths/weaknesses, agreements, and recommendations.

2. Refinement of the video meeting bot (Sembly)
I will add a script for the second meeting, a separate checklist, and correctly separate the logic so that evaluations do not overlap.

Architecture:
… Ringostat / Sembly → n8n → transcription → AI analysis → Kommo + table + logging.

I work systematically: I think through the logic, error handling, and scaling to ensure the solution is stable and not "temporarily working."
You can familiarize yourself with the case example at the following link: https://content.freelancehunt.com/snippet/81a17001-3e9a-42b9-85af-146cbd9a6688/1951609/AI-система%20анализа%20звонков%20из%20СРМ%20с%20отчётами%20и%20дашбордом%20в%20Google%20Sheets.pdf
Please let me know if there are already access to the API and ready checklists for both types of meetings?

Show original

Translate