Nikolay Sekachev Agentic AI: Business Process Automation
A complete implementation track for building AI-powered business workflows: from LLM fundamentals to SQL, deployment, and multi-agent systems.
Practical program (120 academic hours) 6 weeks, blended format (offline + online + independent work) Business owners, operations leaders, analysts, project managers, and technical specialists
Course Positioning
This course is designed for teams that want measurable operational results, not just AI demos.
You move from basic model literacy to production-grade automation of real business processes.
Learning Outcomes
- Design AI workflows for repetitive and multi-step business tasks.
- Connect LLMs with tools, databases, and external services safely.
- Build and deploy practical AI assistants and agent systems.
- Introduce human-in-the-loop controls for critical actions.
- Reduce manual workload and increase process speed and consistency.
Module Structure
Module 1. AI Foundations and Model Ecosystem
- How neural networks work in practical terms.
- Model landscape: proprietary and open models.
- Cost structure, compute, and why model choice matters for business.
Module 2. LLM Mechanics, API, and Prompting
- Client-server architecture and API interaction patterns.
- JSON outputs, token logic, and model parameter control.
- Prompting patterns for predictable business outputs.
Module 3. Agent Systems and Orchestration
- Agent anatomy: memory, tools, planning, execution.
- ReAct, planning, reflection, and task decomposition.
- MCP and dynamic context for tool-enabled workflows.
Module 4. Production Architecture for AI Services
- Frontend/backend responsibility split.
- Data flow design and integration points.
- MVP architecture for internal tools and client-facing services.
Module 5. Deployment and DevOps
- Linux terminal essentials for operators and builders.
- DNS, reverse proxy, HTTPS, and service reliability.
- Deploying frontend and backend with basic observability.
Module 6. Vision and OCR for Operations
- Document understanding and extraction pipelines.
- OCR for invoices, forms, and internal records.
- Practical multimodal use cases for operations teams.
Module 7. Voice and Audio Automation
- Speech-to-text and text-to-speech in business workflows.
- Voice interfaces and call-related automation scenarios.
- Audio pipelines for support and internal productivity.
Module 8. Databases and SQL
- Relational schema design for business entities.
- SQL querying and data quality workflows.
- AI-assisted SQL for reporting and process control.
Module 9. Multi-Agent Infrastructure
- Coordinating specialized agents in one process.
- Secure tool access, permissions, and approvals.
- Production setup patterns for resilient business automation.
Final Project
Each participant builds and presents an MVP for a real business process:
- automation scenario definition,
- implementation architecture,
- working prototype with at least three technologies from the course.