Дмитро

DMYTRO KHODOV
SUMMARY
AI Engineer specializing in LLM-powered automation, multi-agent workflows, RAG
systems, and business process optimization. Hands-on experience building
production-oriented AI pipelines with n8n, FastAPI, PostgreSQL, Docker, Google
Drive, Bitrix24, YouTube Data API, Gemini, OpenAI, and dashboard interfaces.
Focused on turning AI capabilities into reliable workflow systems with measurable
business impact, including manager workload reduction, automated content
operations, and multilingual AI outputs.

WORK EXPERIENCE
AI Automation Engineer - 4Limes
[открыть контакты](см. выше в блоке «контактная информация») May 2026 - Present
Building production AI automation systems for business workflow optimization
Delivered end-to-end process automation with measurable operational impact
[открыть контакты](см. выше в блоке «контактная информация») Developed and integrated scalable AI-powered workflow solutions
Prompt Engineer - Collective Forecast
Linkedin July 2025 - Present
Designed and implemented agent-based systems for financial and crypto market
analysis
Telegram Developed and optimized prompts for autonomous agents performing data
analysis and reasoning tasks
Built multi-agent workflows for automated data processing and insight generation
HARD SKILLS Improved output quality and reliability of agent systems through iterative prompt
and logic refinement
Python (Pandas), SQL
AI Systems Integration PROJECTS
Data Analysis & Data Processing
YouTube AI Automation Platform
Prompt Engineering
Designed and built an end-to-end AI automation pipeline for YouTube production
API Integration
operations using n8n, FastAPI, Docker, PostgreSQL/PostgREST, and a custom
RAG Systems & Vector Databases
dashboard
LLM Application Development
Implemented Google Drive intake for long videos and Shorts, with queued
Multi-Agent Workflow Design
processing, retry logic, stale-job recovery, and workflow status tracking
Production AI Pipelines
Integrated 3 transcription providers with provider selection, fallback logic,
TOOLS chunking, and timed transcript generation
Developed AI workflows for thumbnail creative briefs, YouTube SEO metadata,
OpenAI API, YouTube Data API, Bitrix24 API
Community posts, and structured multilingual content outputs
Google BigQuery, Google Sheets
Integrated Bitrix24 task creation for production handoff, including project-specific
Tableau, Power BI
manager/producer assignment logic
Docker, PostgreSQL
Built YouTube publishing automation for video uploads, metadata submission,
LangGraph, LangChain
subtitles, OAuth channel connection, and upload status tracking
Streamlit, Railway
Created dashboard workflows for project management, queue administration,
Qdrant Cloud, Gemini Embeddings
prompt presets, and operational monitoring
Git / GitHub
Reduced estimated manager workload from 54h to ~9h per client/month and
n8n, FastAPI
increased potential manager capacity from 3-4 to 8-10 clients

SOFT SKILLS Stock Analyst Agent | Live Demo
Problem-solving Built an agentic stock analysis pipeline using LangGraph and Groq
Analytical thinking Developed 3 data tools for real-time price, fundamental, and news data via yfinance
Communication Designed a 9-section structured analysis prompt with multilingual output
Attention to details Implemented a deterministic Portfolio Builder with risk, horizon, style, ETF
Adaptability preference, sector tilt, and allocation logic
Integrated RAG via Qdrant Cloud and Gemini Embeddings for product docs,
LANGUAGES methodology, and company KB
English - B2 Built S&P 500 company KB ingestion: markdown profiles, chunking, embedding,
Ukrainian - native vector upsert into Qdrant
Russian - fluent Added product knowledge agent for architecture, setup, and deployment Q&A
Implemented safe fallback for LLM rate-limit errors with local markdown summaries
EDUCATION
Amazon EU Performance Dashboard | Live Demo
IT School “GoIT”
Built a Next.js + Python/pandas dashboard for Amazon EU sales, profit, margin, PPC,
Data Analytics (2024) and SKU analysis.
Horlivka Institute of Foreign Languages Automated messy CSV processing into clean JSON analytics with country, portfolio,
Bachelor of English (2006-2010) and product-level metrics.
Integrated Groq AI for executive insights, priority SKU detection, and Q&A over
processed business data.