Work Experience
AI & ML Engineer | Back-End Developer (n8n Automations & Agents)
Duration: 0.5 years
Summary:- An n8n AI-powered chatbot developed as an internal company tool for lead generation and customer engagement
- The system enables automated communication with potential clients through WhatsApp, captures and qualifies leads, and synchronizes data with CRM systems
- The architecture is built around Retrieval-Augmented Generation (RAG) for knowledge-based responses and n8n agentic workflows that orchestrate AI interactions, automation pipelines, and integrations with enterprise tools such as Salesforce and Google Workspace
Responsibilities:- Designed chat logic and conversational flow for an AI-powered chatbot.
- Developed automated workflows in n8n to orchestrate AI conversations, data enrichment, and CRM-related processes.
- Built an AI agent with access to multiple tools, including checking available time slots, booking appointments, and saving leads to Salesforce.
- Implemented WhatsApp integration to enable real-time conversations with potential clients and automated lead capture.
- Designed a RAG architecture using Pinecone vector search to retrieve relevant company information and deliver context-aware responses.
- Integrated Salesforce CRM with n8n workflows to automatically create leads, update customer records, and trigger sales processes.
- Configured Supabase as a persistent memory and user data store to maintain conversation context across sessions.
- Applied prompt engineering strategies to guide LLM behavior for structured outputs, lead qualification, and reliable tool usage.
- Built error handling workflows in n8n to ensure system reliability and graceful failure recovery.
Technologies: Python, JavaScript, n8n, Supabase, Streamlit, GPT, Pinecone, Airtable, Salesforce, Google Workspace (Gmail, Google Drive, Google Calendar), Microsoft Outlook & Calendar, WhatsApp, Prompt Engineering & Evaluation (Promptfoo, LangSmith)
AI/ML Engineer | Backend Developer (AI-Driven Discord Bot)
Duration: 0.5 years
Summary:- An automated moderation system for Discord designed to facilitate structured debates on various topics
- The project features complex multi-user interaction flows, real-time voice message processing (Speech-to-Text), and an automated judging/moderation engine that ensures compliance with debate rules and etiquette
Responsibilities:- Designed and implemented backend services for a Discord bot supporting structured multi-user debates and automated moderation.
- Developed streaming TTS, STT, and TTT functionality.
- Built an agentic workflow with custom tool integrations, including automated moderation and dynamic topic management.
- Developed logic for post-session analytics generation, including conversation summaries, debate scoring, sub-topic detection, per-user activity reports, and conversation tips.
- Developed NLP-based components for detecting rule violations, offensive language, and off-topic responses.
- Optimized inference pipelines for low-latency responses during live debate sessions.
Technologies: Python, PostgreSQL, LangChain / LangSmith, groq, ElevenLabs, Whisper, DiscordAPI, WebSockets, OpenAI Realtime API
AI & ML Engineer | Back-End Developer (Mental health chatbot)
Duration: 0.5 years
Summary:- Chatbot designed to help people with their mental problems
- It consists of a well-trained LLM model that is secure
- The bot can interact with users verbally and by text messages in real time, has memory of previous conversations, simulates natural conversation, and offers guided therapeutic exercises
Responsibilities:- Performed back-end development.
- Conducted prompt engineering.
- Developed scripts.
- Scraped and processed documents.
- Developed streaming TTS, STT, and TTT functionality.
- Attached memory to the LLM.
- Developed a simple front-end for the chatbot.
Technologies: Python, FastAPI, React, PostgreSQL, Docker, LangChain / LangSmith, LLaMa, Whisper, ElevenLabs, AWS Bedrock, AWS SageMaker, Selenium, BeautifulSoup (bs4), AssemblyAI API, pyannote, YouTubeAPI
AI/ML Engineer | Prompt Engineer (Website generator)
Duration: 1 year
Summary:- A service that automates the creation of SEO-optimized websites featuring a large number of interconnected pages
- Each website follows a tree-like structure, where pages on a specific topic link to more detailed pages within the same theme
Responsibilities:- Structured page elements.
- Conducted prompt engineering (Few-Shot, Tree-of-Thought, Self-Consistency, Chain-of-Thought).
- Generated media content.
- Performed keyword research.
- Generated tree structures for website organization.
- Conducted content humanization research.
- Performed unit testing.
- Scraped websites for data and content.
Technologies: FastAPI, SQLAlchemy, LangChain / LangSmith, Instructor, treelib, OpenAI API, DALL-E, HeyGen, ElevenLabs, Google Ads API
AI/ML Engineer | Back-End Developer (Data Analytics Chatbot)
Duration: 0.5 years
Summary:- A chatbot designed to leverage LLMs for language understanding, analytical understanding, logical reasoning, and analytical computations
- It works with various dataset formats, generates and executes SQL/Python code based on user requests, solves logic reasoning tasks, visualizes data, and recommends visualizations for user-specific needs
Responsibilities:- Performed back-end development.
- Trained and developed ML tools for the chatbot.
- Conducted prompt engineering.
- Used LLMs (Claude 3.5 and GPT–4/4o) to analyze and orchestrate chatbot agent behavior.
- Created analytical and mathematical tools to extend LLM capabilities.
- Integrated various APIs.
- Implemented and maintained a scalable and robust infrastructure using AWS/GCP.
- Ensured high levels of security and data protection.
- Developed and maintained API documentation.
- Implemented new features.
- Participated in UI development.
- Conducted code reviews.
Technologies: Python, pandas, PostgreSQL, LangChain, transformers, LIDA, PAL, matplotlib / Plotly / ggplot, Streamlit, OpenAI API (GPT-3.5/4/4o), Anthropic API (Claude 3.5), AWS (DynamoDB, EC2), GCP
NLP Engineer | Back-End Developer (Sales bot)
Duration: 1 year
Summary:- A software tool that uses a filter system to select applications that do not need to be processed by sales managers
- The system uses criteria such as customer interests, geographic location, technical skills, and other factors to accurately select potential customers
Responsibilities:- Created models for content generation.
- Utilized Selenium for search automation.
- Performed unit testing.
- Built microservices.
- Designed database architecture.
- Improved the matching system.
Technologies: Python, FastAPI, Selenium, BeautifulSoup, langdetect, Redis, DynamoDB, JWT, Pytest, Docker, AWS (S3, EC2, CDK), Chrome, OpenAI ChatGPT
Education
- Bachelor’s degree in computer linguistics
Taras Shevchenko National University of Kyiv
Certification
- DeepLearning.AI TensorFlow Developer
Coursera | DeepLearning.AI
- NLP – Natural Language Processing with Python
Udemy
- Multi-Agent Systems with LangGraph
Datacamp
- Analyze Speech and Language with Google APIs Skill Badge
Google Cloud
- Using the Google Cloud Speech API
Google Cloud
Upstaffer since May, 2026
