Oleksii Yakubets
Offer Oleksii work on your next project.
Private message
Offer work
Lvov,
Ukraine
7 hours 7 minutes back
available for hire
1 proposal made
age 42 years
on the service 10 months 8 days
Rating
Successful projects
Average rating
Language proficiency level
Українська: fluent
Русский: fluent
English: pre-intermediate
Polski: elementary
CV
AI Integrator | Business Process Automation | No-Code Expert
I help businesses transform their operations, save time, and enhance customer interaction through intelligent automation. I specialize in AI, chatbots, and no-code solutions, implementing customized systems that eliminate routine tasks and significantly boost business efficiency.
What I Do for Clients:
AI Agent & Bot Development: Automating interactions and optimizing business processes.
AI-Powered Content Generation: Setting up systems for automated marketing content creation (ChatGPT, OpenAI, Claude).
Business Process Automation: Optimizing marketing, CRM, Google Sheets, and workflows.
Integrations & No-Code Development: Seamlessly connecting platforms via Make.com (formerly Integromat), Zapier, Airtable, building Telegram bots, and integrating Calendly.
My Key Skills:
AI: ChatGPT, OpenAI, Claude, AI Agents, AI Bots, Content Generation.
No-Code: Make.com (Integromat), Zapier, Airtable.
Automation: Marketing, CRM, Google Sheets, Workflows.
Integrations: CRM systems, Social Media, Email, Messengers.
Let's collaborate!
Write or call to discuss your projects and automation possibilities.
Skills and abilities
Programming
Services
Portfolio
- All works 13
- Content Management 1
- Transcribing 1
- Social Media Marketing (SMM) 1
- Data Parsing 2
- AI & Machine Learning 3
- Data Processing 2
- Bot Development 1
- AI Content Creation 1
-
Lead generation in Telegram: From "manual" search to AI sales
AI & Machine Learning
Pain points of your business (What hinders growth?)
Lost time: Managers spend hours monitoring dozens of chats and groups in search of targeted inquiries, yet still miss 50% of leads.
… Slow response: By the time a person sees a message in the group and writes to the client, competitors have already closed the deal. Speed in Telegram is everything.
Bot limitations: Regular bots do not see messages in closed chats and cannot write to clients first.
Human factor: Managers get tired, make mistakes in names, or forget to respond on time.
Project: Intelligent "Radar" based on Userbot + AI
Developed a system that works like an "invisible manager" that never sleeps.
How it works:
Scanning (Radar): Userbot, disguised as a real account, is present in all important chats for you (even closed ones). It instantly reads every message.
Smart filtering (n8n + AI): The system does not just search for words; it understands the context. Thanks to artificial intelligence, the bot distinguishes a real client inquiry from spam or advertising.
Instant contact (Outbound): As soon as a target lead is found, the system automatically sends a personalized message to the client in private chats.
Personalization: The bot addresses by name and mentions the specific user inquiry, creating the impression of a real live dialogue.
I develop bots for your specific tasks — from smart lead generation systems to complex AI assistants.
-
Transcription, Analysis and Reports (n8n + Gemini AI)
Transcribing
A fully autonomous script has been developed in n8n, which automates the routine process of processing video and audio recordings of work meetings, interviews, or Zoom calls. The system converts the "raw" recording into a structured business report without human involvement.
How the system works:
Auto-upload: The script automatically retrieves new audio files from Google Drive.
… Intelligent transcription (Gemini 1.5 Flash): Using AI, it performs accurate speech-to-text conversion (in Ukrainian) with automatic speaker separation.
Deep analysis (AI Business Analyst): A specially configured model analyzes the dialogue and extracts only the "dry residue":
The main goal of the meeting and the client's key issue.
The technical stack and existing systems of the client.
Priority tasks and critical requirements for the solution.
Action Items: a clear list of next steps (who, what, and when needs to be done).
Documentation and notification: The system creates a document in Google Docs, where the full text of the dialogue and the analytical report are stored.
A brief summary is instantly sent to Telegram for prompt team notification.
Business outcome:
Time savings: Instead of hours of manual work — a ready report in just a few minutes.
Control of agreements: No important detail or assignment from the conversation will be lost.
Cost optimization: The use of the Gemini Flash model allows for high-quality analysis at minimal API request costs.
-
Autonomous AI Editing: Full Cycle SMM Automation (n8n + AI)
Content Management
A unified integrated ecosystem has been developed for the full cycle of content production: from autonomous market research to automatic publication on social media. The project is implemented as a complex multi-threaded scenario in n8n, ensuring stable operation without the need for third-party interfaces.
Deep Research through Firecrawl: The system automatically scans up to 8 target web resources on the chosen topic. This allows for obtaining relevant facts and trends that become the basis for generating relevant publications.
Strategic generation (Gemini Flash): Contextual continuity logic is used — each new post takes into account the content of the previous one, forming a cohesive story for the audience. Individual prompts have been developed for each day of the week, adapting the style and presentation of the content.
Multimedia support (Vertex AI): Visual content for each publication is automatically created using Vertex AI, allowing for the generation of unique images that accurately match the context of the article.
… Scalability and stability with Baserow: A transition from Airtable to Baserow as the main database has been made. This solution was critically necessary, as during intensive work of AI agents in conjunction with external services, Airtable completely exhausted monthly limits on API calls, leading to system shutdown. The use of Baserow has ensured uninterrupted processing of large data arrays.
Publication channels: LinkedIn, Facebook, Telegram.
Result: A full-fledged "AI editorial" has been created, which independently finds information, develops a weekly content strategy, and distributes it across multiple platforms simultaneously. The system is fully autonomous and resilient to loads, allowing for the scaling of content marketing without increasing staff or budget for API limits.
-
n8n Content Factory. Airtable, Softr
An automated system for generating and publishing content for social media has been created based on n8n, Airtable, and Softr. The system consists of two main workflows that operate sequentially for the complete cycle of content creation and publication. The first scenario automatically generates content based on a specified topic: an AI agent creates an article, and the system generates a corresponding image and video related to the publication theme. All generated content is stored in Airtable with the status "Ready for Review." The second scenario is activated after human review: when a user reviews the content through the Softr interface and changes the status to "Approved," the system automatically publishes the article along with media files on LinkedIn and Facebook at the specified time. A fully functional user-friendly interface has been developed on Softr, which provides the user with complete control over the system: viewing generated content, changing publication statuses, editing generation parameters, managing the publication schedule, and configuring output data for the AI agent. The technical stack includes: n8n for orchestrating workflows, Airtable as the central database with relational links, Softr for the frontend interface, AI APIs for generating text and media content, LinkedIn and Facebook APIs for automatic publishing. Automation covers: AI generation of articles on a specified topic considering context and style, automatic creation of visual content (images and videos), a two-step approval workflow with editing capabilities, smart scheduling of publications considering time zones, and real-time synchronization between Airtable and Softr. The result: an 80% reduction in content creation time, the ability for one person to manage 50+ publications per month, complete automation of the process from generation to publication, a structured content calendar with visual representation, and maintaining content quality due to mandatory human review before publication. The system is scalable for marketing agencies, content creators, and small businesses.
-
AI-Agent n8n
AI & Machine Learning
An AI-Agent Telegram bot has been created for automating consultations and booking meetings. The system is built on n8n with integration of Claude AI Agent, Google Calendar, and Google Sheets. The bot independently qualifies leads, adapts communication to the specific niche of the client (small business, corporations, NGOs), offers relevant solutions, and conducts dialogue until booking a 20-minute consultation. The AI Agent works with a detailed prompt (9500+ tokens) that covers response scenarios, objection handling, tone rules, and dialogue structure. Automation includes: checking available slots in the calendar in real-time, collecting contact information (name, email, phone, business direction), two-step confirmation before creating an event, automatic recording to Google Sheets, and the ability to cancel meetings through the bot. Result: 24/7 availability, reduced manual work in processing requests, a structured lead database, and increased conversion due to a personalized AI approach.
-
Migration from Make.com to n8n
Data Parsing
The initial setup of the project on Make.com successfully established a data pipeline but resulted in unsustainable operational expenses (OPEX) for over 100,000 profiles due to high operational costs. An economically efficient and scalable replacement was urgently needed. All automation was strategically migrated to a self-hosted instance of n8n. This transition provided fixed hosting costs and eliminated operational fees. The n8n code was configured to calculate complex metrics (ER_AVG, ER_MED). The high-cost process on Make.com was transformed into an extremely cost-effective and scalable n8n solution, achieving a critical reduction in operational expenses while providing AI-enhanced capabilities.
-
Verification of profiles on Instagram and data enrichment based on AI
Data Parsing
Created a comprehensive automation system for the client to check large lists of Instagram profiles. Selected profiles based on complex criteria, prevented duplication, and effectively collected deep metrics. Used Make.com for conditional filtering and duplicate checking, then analyzed the data using Apify. The main task was to optimize the calculations of the metrics ER_AVG and ER_MED, using minimal Make.com operations to save costs. A fully automated, cost-effective data transfer channel was created, delivering targeted, enriched profiles directly to the client's database. This system replaced manual verification.
-
Make.com Automated Weekly KPI Report for Marketing Agencies
Data Processing
#Make.com was used to develop and configure an automated weekly reporting process for a marketing agency. Using conditional, approximate data, I built a working template to collect, process, and prepare KPI reports in PDF format via Google Slides. The system automatically sends reports to the client via Gmail. The client independently tracks their performance, but this project demonstrates a scalable automation solution ready for real data integration and delivers professional reports.
-
Make.com Automated LinkedIn Content Generation & Publishing Syst
Social Media Marketing (SMM)
This Make.com project automated LinkedIn profile content, ensuring regular updates. The system automates content creation from topic input to publication. It generates text and images using advanced AI tools based on a given theme. For quick feedback, a Telegram notification is sent for article structure review. The Make.com scenario operates automatically Monday-Friday, ensuring consistent content flow.
-
Voiceflow AI Chatbot for Fitness Studio
Bot Development
#An AI chatbot on Voiceflow was developed for a fitness studio to answer frequent client questions, significantly reducing admin workload and freeing up their time. The bot was configured to generate new contacts directly into Google Sheets via a Make.com webhook integration. This ensured instant, accurate answers, improving response times. New leads are now automatically captured and organized, optimizing lead generation. The solution boosted operational efficiency and ensured no contact opportunities were missed, allowing the studio to focus more on client service and growth.
-
AI-Powered Content Automation & Distribution for Personal Brands
Data Processing
#Developed an innovative AI content automation system specifically for personal brands, focusing on process optimization and social media content distribution (Telegram, Facebook, Instagram, LinkedIn). The project involved automatic video selection and transcription. Following this, OpenAI generated unique articles for various social media platforms based on the video text. The system also automatically created corresponding images for each platform and ensured seamless content publication. This solution significantly accelerated the content creation cycle, freeing up time for strategic brand development. Tools: Replicate, Dumpling AI.
-
Make.com + OpenAI: Auto-Generate Blog Ideas in Sheets/Airtable
AI Content Creation
#A Make.com automation was deployed to generate 5 weekly blog article ideas using OpenAI. A custom table in Google Sheets or Airtable (with columns for Keyword, Title, Description, Format, CTA) provides structured output. Running weekly, it extracts keywords, optimizes prompts, and records AI-generated ideas. This solution ensures a steady stream of content concepts, significantly easing content creation. It saves time, boosts content quality/diversity, and drives audience engagement/growth.
-
Calendly Automation: Efficient Booking Management
AI & Machine Learning
#A Make.com automation and Telegram bot were developed to send real-time notifications for new Calendly bookings. Notifications include the client's name, email, appointment date, comments, and buttons to cancel or reschedule.
Implementing this automation allowed the client to respond quickly to new bookings, significantly boosting business process efficiency. Instant communication enabled real-time availability analysis, which, in turn, elevated the overall service level and customer satisfaction.
Activity
| Latest proposals 10 | Budget | Added | Deadlines | Proposal | |
|---|---|---|---|---|---|
|
Looking for an n8n specialist
45 USD
|
|||||
|
Bot for parsing and posting
23 USD
|
|||||
|
AI agent for booking service (on n8n or similar automation system)
259 USD
|
|||||
|
Create a connection between Telegram bot and AI agent
140 USD
|
|||||
|
AI automation, inventory accounting in n8n
162 USD
|
|||||
|
Automation of business processes using AI
349 USD
|
|||||
|
PDF generation (Airtable - PDFMonkey)
20 USD
|
|||||
|
AI Assistant and AI Agent Project Manager
20 USD
|
|||||
|
Integration between respond.oi and crm through n8n
90 USD
|
|||||
|
Telegram bot
101 USD
|
