Maksims Burcenko — AI project manager

Ask this page

You have a question about my background. Instead of scrolling for it, ask. This box runs on a language model that can see my CV and nothing else — the same kind of system I build every day. It will refuse to answer anything the CV does not cover.

In numbers

20+years of professional experience—banking, municipal, and public administration

6years leading Information and Communication Technology (ICT) projects and development teams

9person Information Technology (IT) department, which I led at the National Health Service

3other institutions—the Prosecutor's Office, the Corruption Prevention and Combating Bureau (KNAB), and the Court Administration—within a working group that I initiated, officially established, and led

4years of fully remote work—from 2021 to 2025, I managed national-scale ICT programs without physical office presence

3countries outside Latvia where I have lived and worked, communicating daily in English

8 000+entries written into the long-term memory of the AI agent system I built and have used in daily work since 2025

14routines that run on schedule without me present—they prepare documents, monitor sources and report back

These figures reflect scale—duration and volume. They are verifiable facts. I do not make unsubstantiated claims about the exact percentage by which AI has improved my efficiency, as there are no objective baseline measurements from before the system's implementation.

Summary

I am a project manager. For twenty years, my primary objective has been identifying dysfunctional processes, determining their root causes, and enacting necessary changes by engaging individuals who are not my direct subordinates. Since 2025, I have developed and integrated my own AI agent system into my daily work. This is not a side project; it is the exact same management work, executed with higher-efficiency tools. Because of this, my approach dictates analyzing the actual workflows first, and only then selecting the appropriate technological solution.

My hands-on experience is in public administration, municipalities and the financial sector — environments where one process touches thousands of people and a mistake is expensive. When I join a new company, I trace the process end to end first, and only then say how it should run.

AI implementation does not fail because the tools are inadequate. It fails because employees do not utilize them in practice.

Current practice · 2025 — present

AI agents and process automation

I operate independently and without direct supervision—this alone demonstrates my capacity to effectively manage remote work.

Agent system in daily work

Task automation, document preparation, monitoring, report generation, and workflow handovers between stages. I architected the entire system myself, maintain it independently, and assume full end-to-end responsibility.

Skill files

I independently develop and refine the instructions that dictate agent behavior, actively rewriting them when real-world usage exposes flaws in the initial directives.

Large language models via direct system interfaces (API)

"Claude" and "Gemini". I select the model for each task individually, optimizing the cost-to-quality ratio while accounting for potential model changes or downtime, rather than relying on a single solution's permanent availability.

Data location

Data is stored locally on my hardware—both the database and agent memory. External data transmission is always a conscious, controlled decision, never an automated habit.

System integration, known as Model Context Protocol (MCP) servers, and tool access

I ensure agents connect securely to necessary systems—browser control, databases, scheduled tasks—and I independently identify and resolve integration failures.

Written by default

Instruction (prompt) and Standard Operating Procedure (SOP) engineering to guarantee workflow repeatability, independent of the system's creator.

Testing after every change

Following any alteration to a model, instruction, or data structure, I re-execute system tests and measure the results before confirming a process improvement.

Governance

Three levels established before agent deployment

An agent without strictly defined boundaries is not an efficiency tool—it is a disaster waiting to happen. Boundaries are architected before the process begins, not as a reaction to an error.

Permitted without questions

Tools the agent is authorized to use autonomously—reading, data retrieval, text generation, and local processing. These are reversible actions executed strictly within my personal infrastructure.

Requires human approval

Any action that exits the local environment, generates costs, or is irreversible—publishing, sending emails, external system calls, or data deletion. The agent prepares the content; a human authorizes the execution.

Forbidden

Access credentials, financial transactions, irreversible actions, and anything the user has not explicitly requested. During runtime, the agent is restricted from granting itself exception privileges.

A solution that functioned flawlessly last month can silently break after a new model version is released. Therefore, technical tests are repeated after every modification—such system defects rarely trigger alarms; instead, they silently degrade the process, necessitating proactive monitoring.

How I got here

AI with concrete facts, no embellishment

2024Implemented AI tools into daily workflows at the Competition Council—document drafting, summarization, and data analysis. Graduated from the School of Public Administration program "Artificial Intelligence Methods for Daily Work."
2025Initiated the independent development and practical deployment of my own AI agent architecture, moving away from standard third-party commercial products.
2026Earned a certificate in the program "Use of Artificial Intelligence Tools and Agents in Daily Work and Business" (timefor.ai, 29 hours).

I began my professional career in the banking sector, making the strict requirements of heavily regulated industries standard practice for me. Consequently, I fully understand that actual AI implementation in enterprise organizations moves significantly slower and more cautiously than it appears in technology demonstrations.

Experience

IT Project Manager

Competition Council, Riga

2021 — 2025

Worked remotely for the entire four-year tenure. This format began due to pandemic restrictions and was maintained permanently. Everything detailed below—program development, budget reporting, working group coordination—was accomplished remotely.
Managed ICT development programs from business requirement definition through implementation and testing, coordinating 10–12 stakeholders: business process managers, analysts, and developers.
Proposed the implementation of a national electronic platform for court cases (E-case) at the institution and drove its adoption.
Initiated, officially established, and led an inter-institutional working group (Prosecutor's Office, Corruption Prevention and Combating Bureau (KNAB), Court Administration, and the Competition Council). Participants were officially delegated by their respective institutions. The decisions made by the group served as the foundation for developing the Competition Council's E-case information system.
Maintained focus on priorities during situations with conflicting stakeholder demands; consistently reported to executive management on project progress, budget execution, and potential risks.
Drafted an application for a European Union (EU) support program and organized an international experience exchange visit to Finland.
Successfully delivered five ICT initiatives from business requirements to deployment: cartel screening tool roadmap, E-case implementation, survey module development (including market research), merger process digitalization, and the introduction of a data visualization platform (Tableau). For each project, I managed technical specifications, working groups, procurement, and on-time delivery.
Implemented the Tableau data visualization platform: reports were adopted by three departments. The reporting cycle, which previously required three to four days of manual data processing, was reduced to a single day.
Deployed AI assistant (Copilot) licenses at the executive management level (4 licenses)—this was the first AI tool integrated into the daily work of the institution's leadership, for which I also provided user training and technical support.

Head of IT Project Development Division

National Health Service, Riga

2018 — 2019

Led a team of nine IT specialists responsible for developing the national E-Health information system.
Oversaw the ICT project portfolio; planned resources, implementation methodologies, and inter-institutional coordination.

Warehouse Supervisor; Production Operator

Templine (Kettering); Rigid Containers Ltd (Desborough), United Kingdom

2016 — 2018

Worked in the manufacturing sector in England, concurrently earning an Institute of Leadership & Management (ILM) Level 3 qualification as a "Team Leader / Manager" (EQV (UK) Ltd) during evening studies.
Within months of returning to Latvia, I was managing a nine-person IT department. I do not hide this period, as it explicitly demonstrates my capacity to rapidly adapt and master new industries from scratch.

Co-founder, Board Member, Head of Marketing and Sales

SIA "Psaiho.lv", Riga

2011 — 2015

Co-founded the company and served as a board member and co-owner.
Directed marketing and sales operations: acquiring clients and ensuring high-quality service delivery.

Director, Board Member

PSIA "Jūrmalas Attīstības Projekti", Jūrmala

2010 — 2014

Director of a municipally-owned enterprise: held full accountability to the owner for the company's operational activities, budget planning, and financial reporting.
Operated under municipal oversight—justifying decisions and defending them within a political environment.

Project Manager

SIA "Sintegra", Microsoft Gold Partner, Riga

2008 — 2009

Managed IT solution implementation projects in the private technology sector: analyzed client requirements, planned deployment phases, and facilitated stakeholder communication.

Securities and Foreign Exchange Transactions Specialist; Customer Service

AS "SEB banka", Riga

2005 — 2008

Processed securities and foreign exchange transactions within a regulated financial environment, specifically in the support function division (back office).
Initially performed customer service and cashier duties—spending my first three years in direct client engagement.

How I work

Four years remote in a state institution

Worked entirely remotely at the Competition Council (2021–2025). All deliverables—software platforms, budget reports, inter-institutional working groups—were achieved remotely.

Remote and unsupervised

Since 2025, I have independently managed my own practice: defining priorities, setting deadlines, and architecting systems. My workflow requires no external micromanagement.

English as a working language

I have lived and worked in three countries outside Latvia—the United States (2003), New Zealand (2010), and the United Kingdom (2016–2018). My ILM Level 3 qualification was earned and assessed entirely in English.

Leading beyond direct reports

Initiated and led an inter-institutional working group with officially delegated participants, coordinating collaboration across ministries, judicial bodies and the prosecution service.

Corporate contract and self-employment

I have executive experience as a company director and board member; I am fully equipped to engage in B2B (business-to-business) contracting, independently managing invoicing and reporting.

Time zone overlap

Based in Baldone, Latvia, UTC +2/+3. I am available for afternoon and evening shifts to synchronize with the US East Coast.

Education and certification

Professional Master's Degree in Business Management — Information Systems Management Institute (ISMA), Riga (2008). Bachelor's Degree in Business Management (2006).
Use of Artificial Intelligence Tools and Agents in Daily Work and Business — timefor.ai, 29 hours, certificate (2026).
Principles of the Information Society and E-Governance — Riga Technical University (2025).
Artificial Intelligence Methods for Daily Work — School of Public Administration (2024).
ILM Level 3 qualification, Team Leader / Manager — EQV (UK) Ltd (2017–2018).
Appreciation from the School of Public Administration — for expertise in developing the project management competency framework for public administration (2024).