AI transformation · education · production systems

Make your company capable of building with AI.

I help leadership teams turn AI from scattered experiments into an internal capability: people who understand the economics, systems that work on real data, and an organization that can continue without a permanent external consultant.

Nikolay Sekachev

Education, consulting, and implementation for

Stockholm School of Economics in Riga
OMD Latvia
Baltic Business Club
Eesti Koolitus- ja Konverentsikeskus
Profftech
EVC
Eterna
MixPack
VIP Medicum
Dunven
EDU-DO
Sky Autokool
The operating method

Teach → Build → Transfer

The goal is not dependency on an AI consultant. It is a company that can identify, build, accept, operate, and improve its own AI systems.

01

Teach

Leaders and process owners learn to see where AI changes economics—not merely which buttons to press.

Shared language and a ranked project portfolio
02

Build

A small internal team builds one vertical slice on real data, with a business owner, controls, and an acceptance test.

A working system, not a presentation
03

Transfer

The company receives the repository, runtime, second operator, documentation, measurement loop, and the ability to continue independently.

Capability stays inside the company
Selected systems

Concrete work, with operating logic and evidence

Not generic case-study theatre: what was broken, what the system now does, what is already running, and where human control remains.

BiomotorAI · SalesProduction
Multi-country B2B lead engine
Sales prospecting was fragmented across registries, delivery platforms, maps, and CRM exports.

A scheduled pipeline collects and normalizes hospitality and food-company signals in Estonia, Latvia, and Lithuania, deduplicates against Salesforce, caps uploads, and reports every run.

Production pipeline · multiple public and commercial sources · PostgreSQL · dashboard · backups

BiomotorAI · Commercial qualityWorking MVP
Outbound call intelligence
Managers could not review enough sales calls consistently or turn feedback into repeatable coaching.

A browser call flow records with consent, verifies signed webhooks, transcribes multilingual audio, creates an English working version, scores a checklist, and prepares a manager note.

Working MVP · Twilio · multilingual transcription · coaching score · manager review

BiomotorAI · MarketingProduction
Controlled content operating system
Strategy, source media, localization, production, approvals, and publishing lived in separate tools and chats.

One system connects approved strategy documents and real assets to localized drafts, reusable templates, scoped review, an operator dashboard, and controlled publication preparation.

Production · 65 knowledge documents · 216 media assets · 10 strategy documents · PostgreSQL API

BiomotorAI · AnalyticsValidated pipeline
Explainable customer segmentation
Customer priority depended on one-dimensional turnover reports and individual manager intuition.

A reproducible pipeline combines volume, profitability, and lifetime signals into explicit ABCD scores and aggregates related buyers into a single commercial view.

Reproducible local pipeline · explicit formulas · client aggregation · reviewable outputs

TOOLS_BY · OrganizationOperating program
AI Office: capability, portfolio, and production discipline
Training produced isolated experiments; prototypes lacked owners, acceptance criteria, maintenance, and a route into daily work.

Three learning streams—leaders, managers, and executives—feed one governed project portfolio. Each application receives a business owner, technical operator, repository, runtime, logs, cost visibility, and an acceptance gate.

3 learning streams · 12 leaders · 27 active projects in the current index · 363,654 transcript words studied

Open curriculum

Agentic AI & LLM: learn by building a real system

A 12-module program designed to take an ordinary computer user to an independent builder of agentic systems. The public library currently contains 29 recordings; project work, practice, and preparation complete the academic load.

12 modules29 recordings35:03:00 video120 academic hours
Agentic AI & LLM
Why 35 hours of video becomes 120 academic hours
0135:03 — public recordings; these are part of the contact load
0272 academic / 54 clock hours — contact learning and guided practice
0348 academic / 36 clock hours — independent project work
04120 academic / 90 clock hours total
Point of view

AI creates value only when the operating model changes.

Automating a bad task makes the bad task faster. Transformation begins when real data exposes the bottleneck, a bounded system proves a better path, and roles, decisions, controls, and measurement are redesigned around that new capability.

Teaching and speaking

From executive rooms to hands-on build sessions

Vibe coding: from idea to working product
Talk · DevClub
Vibe coding: from idea to working product
AI for Business Leaders
Executive education · SSE Riga Executive MBA
AI for Business Leaders
Generative AI for business
Workshop · SSE Riga
Generative AI for business
AI for work
Training · Tallinn
AI for work
Advanced AI practicum
Training · Riga
Advanced AI practicum
AI, productivity, and the future of work
Media · ETV+ / Radio
AI, productivity, and the future of work
Hermes · DeepSeek V4 Flash
Ask the public AI advisor
Explore a process, challenge an automation idea, or ask about the course. Public sources only—no access to client data.
0/4000