Personal details
Aleksandr M.
About
Staff Software Development Engineer with over 15+ years of proven success in developing and scaling AI, web, desktop and mobile applications. Has experience with B2B enterprises, startups and large B2C companies. Led implementation of cloud-first event-driven distributed systems. An expert in Python, Java and Go, nowadays delivering primarily by configuring and directing AI coding agents for higher velocity and better quality - picking whichever language the problem calls for rather than the one most practiced, with recent projects shipped in Rust, TypeScript and Python. Aiming to solve business problems by using simple and dedicated strategies.
Hands-on experience in DevOps area with applying CI/CD, GitOps, IaaC, observability tools. Led implementation of applications based on ML, RAG and agentic technologies. Emphasizing the automation of processes, proactive development, commitment to DevOps practices and complete ownership of the system.
Acted as a Team/Tech Lead, managed Agile and Lean teams, optimized processes to drive project success while fostering a collaborative and inclusive environment. Constantly conducting technical and design interviews for SDE, Solution Architect, DevOps and SDET roles. Integrates agentic tooling into engineering workflows as an AI SDLC practice, and maintains open-source developer tooling in this space.
Work Experience
Owned AI SDLC configuration company-wide - integrated agentic coding tools into engineering workflows to raise team velocity and delivery quality, delivering with Claude Code and supporting the company's Anthropic partnership.
• One of the main drivers of the company's spec-driven development ecosystem, built on AWOS - Provectus's open-source Agentic Workflow Operating System. Authored /awos:flow, which generates a project-specific delivery pipeline covering toolchain detection, local review before CI spend, concurrent remote gates, ticket-state lifecycle and post-merge activities. Ran internal demos to grow adoption.
• Rewrote the ai-readiness-audit plugin: the previous implementation was entirely LLM-driven and returned a different score on every run against the same codebase; the rewrite made it idempotent and configurable, so an audit became a reproducible measurement teams could act on.
• Reviewed every change across the AWOS ecosystem other than my own - shaping its conventions, review tone, delivery flow and technical direction.
• Maintained AWOS Recruitment, Provectus's internal skills/agents registry, contributing code review and testing skills/commands into it.
• Designed backend services, ML tasks, Azure infrastructure and inter-component contracts for an AI-driven ERP/document verification application for an enterprise accounting client.
• Fixed chronic code-quality problems for an AI application team at a compliance attestation company (ISO, HIPAA) by introducing AI SDLC practices.
• Enhanced an internal AI-powered Slack bot on AWS by redesigning conversation and transcription flows.
• Conducted technical interviews, primarily for Solution Architect roles.
- Designed and built multi-tenant, cloud-hosted, AI-driven application from scratch.
- Architectured efficient and secure CI/CD pipelines which deliver new features to prod in minutes.
- Optimized web services, OLTP and background jobs for latency, concurrency and low CPU/RAM consumption.
- Integrated with a number of external services to prepare OLAP data storage for AI agents.
- Designed easy-to-use RAG knowledge base updating system with drift detection.
- Executed most of the technical, management, hiring aspects of a fintech startup.