Богдан

BOHDAN PUHACHUK | AI/ML DEVELOPER
Email: [відкрити контакти](див. вище в блоці «контактна інформація») Mobile: [відкрити контакти](див. вище в блоці «контактна інформація») Telegram: @Pah0m4k

Experience

AI Automation & Applied AI Engineer (Python, Computer Vision, ML) Apr. 2024 – Present

• Python Backend (ML/CV/AI Deployment Focus)
o Independently designed and implemented Python backend services using FastAPI as the core deploy -
ment layer for AI solutions, enabling stable inference APIs, authentication, request validation,
logging, and integration with external business systems. Independently designed and evaluated
ideas, offering
the best options instead of impractical solutions.
o Architected backend infrastructure around ML and Computer Vision pipelines, including preprocess
sing layers, inference endpoints, async/background processing, and webhook -based event handling
to support production usage.
o Proposed and delivered API-first system designs that allowed AI models (CV, OCR, LLM-based
components) to be easily integrated into existing business workflows and automation platforms .
• Machine Learning & LLM (Applied Use Cases)
o Designed and implemented applied ML and LLM components inside business automation pipelines,
including structured data extraction, classification-style routing, content processing, and validation
logic for real-world production workflows.
o Built lightweight agent-based flows using Lang Chain and Lang Graph, orchestrating tools, prompts,
and multi-step reasoning logic to automate business tasks without overengineering full agent
platforms.
o Worked with generative AI pipelines and diffusion-based image generation concepts, understanding
inference workflows and integrating generative components into automated content and media
pipelines.
• Computer Vision & OCR Systems
o Proposed and delivered end-to-end Computer Vision solutions for identity verification and document
processing, including YOLO-based detection pipelines and full CV lifecycle: dataset preparation,
annotation, model fine-tuning, evaluation, and deployment into production APIs.
o Designed and implemented a KYC face verification pipeline, including face detection,
preprocessing, matching logic, confidence thresholding, and FastAPI -based deployment for
integration with backend business systems.
o Built OCR-driven financial document processing workflows, including image preprocessing, OCR
extraction, post-processing, validation, and structured output generation for automated backend
pipelines.
• Automation Systems & Business Integrations
o Designed modular automation architectures using n8n, Make, and Flowise, building reusable
workflow components that combined AI services, APIs, databases, and business tools into scalable
automation systems.
o Implemented complex multi-platform integrations using OpenAI API, Gemini API, Meta Graph API,
Telegram Bot API, Stripe API, DocuSign API, Google Cloud API, Slack API, Notion API, and
Airtable API, focusing on secure authentication, reliable data exchange, and production stability.
Skills & Tools

Programming & Backend
• Python (OOP, FastAPI, Flask, AsyncIO, REST APIs, Webhooks); JavaScript (API scripting,
automation logic); API-first backend architecture; async processing; authentication and service
integration.
Machine Learning
• Model fine-tuning; dataset preparation and preprocessing; feature engineering (basic); supervised
learning pipelines; PyTorch; scikit-learn; NumPy; Pandas; model evaluation and inference integration.
Computer Vision
• YOLO (object detection pipelines); CNN-based vision workflows (applied usage); image
preprocessing; dataset annotation and segmentation; OCR pipelines; CV model deployment via API
services.
AI Implementation & Generative AI
• LLM integration (OpenAI API, Gemini API); LangChain, LangGraph (agent orchestration); prompt
engineering; generative content pipelines; diffusion model concepts (applied usage); AI service
selection and integration for business automation use cases.
Automation & Integrations
• n8n; Make; Flowise; workflow orchestration; webhook-based automation; third-party API integrations
(Meta Graph API, Telegram Bot API, Stripe API, DocuSign API, Google Cloud API, Slack API,
Notion API, Airtable API).

Education

o Bachelor of Science in Energy "Renewable Energy Sources"
o Master's Degree in Management "Management of International Business"
o Java Backend Development -- Prog.Academy
o applied AI research & collaboration -- AI Community Program
o n8n: AI Agents, Automation & Voice Systems
o Machine Learning Specialization — DeepLearning.AI & Stanford University
(including Advanced Learning Algorithms, Unsupervised Learning, Recommenders &
Reinforcement Learning)
o Hugging Face — Fine-Tuning and Large Language Models Courses