Skip to content

Alexander Kosykh · Fractional CTO / CIO · Enterprise Architect

IT and AI are a source of competitive advantage, not just "business support"

25+ years designing systems businesses trust with their money: from banking software in 1990s Novosibirsk to solutions for large international companies — Expedia, Coca-Cola, Verizon, GE Capital and others. Built the IT function from scratch three times, ran corporate IT departments and delivered the full lifecycle of multi-year, business-critical IT projects. AI has been a working tool since 2017, not news from a feed.

FIG. 01 — systems map of the practice17 nodes · live
25+years in IT
3IT functions built from scratch
$15M+project budgets under management
100+people in teams I have built
2017AI in production since this year
MicrosoftExpediaSymetraUnisysAxenix / AccentureНЦГИ · MyGeneticsBusiness Strategies FundLikematica.ai

25+ years in IT — from banking software developer in 1990s Novosibirsk to CTO with an international background: Microsoft, Expedia, Accenture, Symetra. Full cycle completed: wrote code for distributed systems, designed Big Data platforms, built the IT function from scratch three times, ran corporate IT departments and delivered the full lifecycle of multi-year, business-critical IT projects.

For me, AI is not news from a feed but a working tool since 2017: bot agents and NLP pipelines at Accenture, ML experiments in insurance, production LLMs today. I work hands-on with a modern stack — LLM, RAG, MCP, self-hosted models (Qwen/DeepSeek, YandexGPT, GigaChat) — and I know how to deploy it safely: 152-FZ, data within the perimeter, integration with 1C and Bitrix24.

Particular expertise — medicine and biotech: three years as CTO of a genetics company, НЦГИ, and the very start of my career at the Institute of Cytology and Genetics, SB RAS. Today I combine two CTO roles in parallel and build my own AI products — so that every day I personally face the limitations of technologies, not just set the direction.

I work in a fractional format — and open to full-time engagements where IT is a source of competitive advantage, not just "business support".
01

Fractional CTO / CIO — retainer

Technology strategy, development, infrastructure, hiring and team growth — on a part-time basis. Covers 80% of a full-time CTO’s tasks without the cost of hiring one.

150–360 k ₽/momonthly · up to 5–6 clients in parallel
02

Express diagnostic

Fast assessment of IT and AI maturity: three main risks, a roadmap with priorities and clear economics. An honest answer on whether you need a CTO at all.

3–5 daysentry into engagement
03

AI audit and pilots

Scenario selection, RAG pipelines, AI agents, self-hosted infrastructure. From prototype to production — with KPIs fixed before the start.

audit → pilot → productioneach stage billed separately
04

Technical due diligence and rescue of stalled projects

Technical expertise for deals and investors: architecture, code, team, tech debt. Bringing stalled projects and teams back to stable operation.

funds · mid-marketexpertise before the deal
05

0→1 launch

Building an IT function from scratch: processes, stack, team. Proven three times — in data centers, biotech and AI products.

processes · stack · teamturnkey infrastructure
06

Sovereign IT / AI

Import substitution, 152-FZ, private LLM infrastructure: your data never leaves your perimeter. Architecture that passes the security review.

152-FZ · self-hostedprivate LLM infrastructure

A full-time CTO in Moscow costs 650 k — 1.3 mln ₽ per month — plus 3–6 months of search, contributions and the risk of a bad hire. A fractional CTO retainer covers 80% of the same tasks for 150–360 k ₽. The 3–5x difference is the main argument for the format.

815 bln ₽the Russian IT outsourcing market — +25% growth in a year: hiring budgets are shifting to external expertise
53%of SMBs already use AI — versus 17% a year earlier; 78% adopt it through external partners
80%of AI pilots in Russia get no continuation — the market pays for reducing the risk of failure

The market pays not for technologies but for reducing the risk of failure — and that is exactly what a senior architect sells.

ParameterFull-time CTOFractional CTO
Monthly cost650 k — 1 mln ₽ (up to 1.3 mln with taxes and bonuses)150–360 k ₽ in 20–40 h/mo packages
Contributions and bonuses+30% insurance contributions after SMB benefits were cancelledNone — contract with a self-employed person / sole proprietor
Time to result3–6 months of search and onboardingStart in 1–2 weeks, monthly contract
Risk of a mistakeHigh: replacing a C-level takes months and costs reputationExit with 30 days’ notice
Annual savings—~5–10 mln ₽/year versus the full-time rate
Step 01Express diagnostic3–5 days · 30–50 k ₽
Step 02AI audit / IT diagnostic2–6 weeks · 100–500 k ₽
Step 03Fractional CTO retainermonthly · 150–360 k ₽

Market price benchmarks, Russia, 2026. Audit cost is often credited toward subsequent implementation.

R1

Three IT functions from scratch

New Hosting Company (specialized data centers), НЦГИ (biotech), Likematica.ai (AI products) — processes, stack, team, infrastructure.

R2

AI in production since 2017

From bot agents and NLP pipelines at Accenture to LLM/RAG/MCP today — continuous practice, not retraining on the hype.

R3

Budgets over $15 mln

Enterprise-level projects under full management; global teams of 20–25 people on three continents.

R4

Windows Live Voicemail at Microsoft

A product taken from early drafts to a production release: a team of 50+ engineers, coordination across the US, Russia and India.

R5

Own AI products with paying pilots

UpStudent — an AI assistant for tutors (NLP); SamBoss — a platform for delivering individual services.

R6

TOGAF 9 + ITIL in one head

Certified Enterprise Architect: architecture and processes are not separate consultants but a single decision-making system.

"AI for work, not for demo"

principle of the practice

We automate processes that actually eat up the client’s time and money — rather than stretching ready-made tools onto someone else’s processes. Against hype, for measurable results: KPIs are fixed before the pilot starts.

2024 — present

Likematica.ainow · parallel project

CTO · AI products
  • Technical vision and delivery of AI solutions for marketing, real estate and the automotive business: automation pipeline architecture, choosing classic vs. AI per task.
  • Adapting GenAI, ML, LLM and MCP to the company’s products; hands-on code on the most complex parts.
  • The principle "AI for work, not for demo": automating processes that actually eat up the client’s time and money.
2024 — present

Business Strategies Fundnow · parallel project

CTO · Moscow
  • Technology strategy for a consulting group: development control (Agile, Scrum, DevOps), cloud and on-premise infrastructure.
  • Selection, motivation and development of key specialists; IT budget planning.
2021 — 2024

MotherHash.io · New Hosting Company

CTO / CIO · data centers
  • IT function and infrastructure of specialized data centers from scratch: monitoring, billing, customer portal, CRM/ERP, information security.
  • As an enterprise architect, orchestrated all systems into a single seamless complex: reliability, integrity, audit, continuous UX.
JavaPythonPrometheus / GrafanaNetBoxBitrix241C ERPAWS · Yandex CloudDockerPostgreSQL
2018 — 2021

MyGenetics · НЦГИ

CTO · biotech
  • Turned a biotech startup at the intersection of science and a direct consumer service into a reliable, scalable business: processes, IT platform, products.
  • Managed a team of genetics scientists, a DNA laboratory, IT infrastructure, PMO and product teams — from scratch.
2017 — 2019

Axenix · formerly Accenture

Head of AI Practice
  • Data Science and AI for large clients: bot agents, office robots, NLP and text/image classification, ML pipelines — with personal involvement in code.
  • Helped build the Data Science practice: team, data architecture, infrastructure (cloud, own data center, hybrid), Data Governance.
WatsonTensorFlowPythonAWS · Azure · GCP · OpenStack
2014 — 2017

Symetra

Enterprise Architect · Bellevue, WA, USA
  • IT roadmap for 1–3 years, control of project compliance with the insurance company’s architectural and security policies.
  • Big Data and EDW, real-time analytics, data integration strategy; ML experiments for underwriting and marketing.
  • Earned and applied the TOGAF 9 and ITIL certifications here.
2009 — 2014

Expedia

Senior Architect · IT Product & Project Manager · Bellevue, WA, USA
  • Data Governance and architecture working groups; upgrading mission-critical databases without downtime for global online travel services.
  • Near-real-time ETL for the enterprise data warehouse: integrating NoSQL (Cassandra, MongoDB, Hadoop, Pig, Hive) with MS SQL and DB2 in a distributed team.
  • Real-time financial integrity audit system; transaction standardization per OAGIS with integration into Oracle EBS.
2008 — 2009

Unisys

Architect · Application Modernization
  • Reusable solution patterns and starter kits on the Microsoft stack: full-fledged prototypes on SQL Server, SharePoint, BizTalk.
2006 — 2007

Microsoft

Senior Program Manager · Redmond, WA, USA
  • Creation and launch of Windows Live Voicemail in partnership with Qwest Communications: all stages — from concept to production release and a series of updates.
  • Scrum master of a 20-engineer team; coordination of teams in the US, Russia and India, including Hotmail, Office, Research, Operations.
1992 — 2006

Early career

consulting · telecom · FMCG · banking software · science
1998–2006
GRT · Stamford, CT, USA
Senior Consultant

Clients: Verizon, GE Finance, Teradyne. Architect of a real-time geospatial data warehouse for a major telecom — 2 years, a team of 25 people.

1996–1998
The Coca-Cola Company · Novosibirsk
Project Manager

All-Russian IT projects; a sales data collection system from dozens of sites with management reporting generation.

1994–1996
Doctor Bit · Novosibirsk
Lead Developer

Banking software for remote client operations; my main build was the remote client for accessing banking systems.

1992–1994
Institute of Cytology and Genetics, SB RAS
Junior Research Scientist

Algorithms for analyzing medical and biological data: cluster and regression analysis, relationship graphs — first hands-on data science experience.

Business strategy & business–IT alignment
TOGAFCOBITITIL 4PMI · PgMPMoP · PRINCE2SAFeScrum · KanbanDevOps
AI and data
LLMRAGMCPNLPML pipelinesYandexGPT · GigaChatself-hosted Qwen / DeepSeekTensorFlowscikit-learnSpark · FlinkHadoopCassandra · MongoDB
Infrastructure and clouds
AWSAzureGoogle CloudYandex CloudDockerPrometheus · Grafanaon-premise · hybrid
Enterprise systems
1C ERPBitrix24CRM · billingOracle EBSOAGIS · ACORD · EDI
Languages & development
PythonJavaKotlinGoTypeScriptRustBash · ZshSQL
Certificates
C-01
TOGAF® 9 CertifiedThe Open Group · enterprise architecture
C-02
ITILIT service management
C-03
MCSDMicrosoft Certified Solution Developer
C-04
Data Science and Data Engineering Bootcampfull-cycle data work
C-05
Elastic Cloud Infrastructure: Containers and Servicescloud infrastructure
C-06
Machine Learningmachine learning
Education

Novosibirsk State University

Faculty of Mechanics and Mathematics · applied mathematics and statistics · 1993 · Novosibirsk

Languages
Russiannative
EnglishC2 · working language for 20+ years in the US and global companies
ItalianB1 · basic working level