Всеволод

Vsevolod Symko
AI Automation Engineer
[open contact info](look above in the "contact info" section) | [open contact info](look above in the "contact info" section) | Telegram: @sevaes | GitHub: github.com/vsevolodshimko

SUMMARY
AI Automation Engineer with hands-on experience designing and shipping end-to-end automation pipelines using n8n
and large language models. Background in product design brings a strong UX-driven approach to workflow architecture
and human-in-the-loop system design. Skilled in building autonomous AI agents, RAG-based knowledge systems,
structured data extraction, and smart escalation logic that reduce manual work and speed up business decisions.

EXPERIENCE
AI Automation Engineer — Inso.code
6 months, through August 2026

• Designed and built end-to-end automation workflows in n8n, integrating LLMs to solve real business problems
across sales, finance, and market intelligence.
• Developed autonomous AI agents for tasks such as market/competitor monitoring, invoice processing, and
customer support, using REST APIs, Cron scheduling, and structured JSON outputs.
• Built RAG (Retrieval-Augmented Generation) systems on vector knowledge bases to enable accurate,
hallucination-free AI responses grounded in company data.
• Applied multimodal LLMs to extract and validate structured data from documents, and wrote JavaScript logic to
enforce data accuracy and business rules.
• Designed Human-in-the-Loop (HITL) and Smart Escalation Routing flows that combined automatic
decision-making with human review for higher-risk or high-value cases.
• Implemented Tool/Function Calling and conversational memory to let AI agents take actions (e.g. registering leads
in a CRM) and maintain context across conversations.

SKILLS
Automation & Orchestration: n8n, Cron Scheduling, REST API / HTTP, Webhooks
AI / LLM: LLM Integration (Text & Vision), Prompt Engineering, RAG, Vector Stores, Embeddings, Tool / Function
Calling, Structured JSON Output
Development: JavaScript (validation & business logic), Data Normalization, Structured JSON Schemas
System Design: Human-in-the-Loop (HITL) Workflows, Smart Escalation Routing, Threat / Risk Scoring, Conversational
Memory
Additional: Product Design background — UX-driven approach to automation and agent workflow design