Summary

Engineering leader, in tech since 2006: AWS and Azure architecture, operations, and applied AI. Currently building an AI SRE product and taking on contract engagements, especially with teams whose AI-accelerated delivery has outrun their understanding of what actually ships.

Engineering leader, in tech since 2006, 15+ years of it building and operating systems, on and off the cloud, with the cloud work primarily on AWS and Azure, across both individual-contributor and management roles. I solve problems, design architectures, build tools, and help the people around me do the same. I care about understanding before automating, and the tools I build are meant to accelerate understanding, not replace it.

Currently building an AI SRE product to keep my hands busy and up to date with the latest tech, and taking on contract engagements alongside it. My sweet spot: teams that went hard on AI in their development process and feel that something is off. Shipping is faster, understanding is thinner, and nobody is sure what actually changed. I help figure that out. Also cloud architecture, operations, and the unglamorous work of making systems boring to run.

Experience

Fractional CTO, OpsWorker (opsworker.ai) 2025 Q4 — Present

AI SRE product: agent architecture and EU-region AWS infrastructure turning Kubernetes alerts into root-cause analyses.

OpsWorker turns Kubernetes alerts into root-cause analyses, on top of the monitoring a team already runs. I lead the technical side: the agent architecture, the AWS infrastructure it runs on (fully inside EU regions), and the engineering decisions behind it, read-only in the cluster by default, human in the loop for judgment. The stack underneath: Amazon Bedrock and Bedrock AgentCore, agents built with the Strands Agents SDK, the Claude and OpenAI APIs, and the Kubernetes API.

Founder, AI and Cloud Architecture Consultant, envimate GmbH 2016 Q3 — Present

Independent AWS & Azure consultancy — cloud architecture, DevOps, applied AI. Siemens, mercateo, and Gulp among the clients.

  • Independent consultancy designing and operating scalable cloud-native systems on AWS and Azure, across DevOps, SRE, automation, and applied AI.
  • The work sits where AI, automation, and cloud architecture meet: infrastructure as code, observability, and self-healing systems. Automate what the team understands, not what looks impressive in a demo.
  • Currently taking on engagements around AI in the development process: teams whose delivery got faster since adopting AI tooling while quality, reviews, or incident response got harder to trust. The work is understanding what actually changed, then fixing that.
  • Selected engagements:
    • Siemens. MindSphere Gateway architecture and code review, with scaling proposals for AWS. Separately, an istio service mesh evaluation with a custom Go performance client across AWS and Google Cloud.
    • mercateo AG. Migration of a large Spring Boot landscape to AWS: CI/CD, infrastructure as code, multi-account deployment pipelines, and cost optimization on ECS, CloudFormation, and Terraform.
    • Gulp. SIEM system on AWS built on Elasticsearch, Kibana, and Beats, plus an automated code review toolchain with SonarQube and CodePipeline.
    • Data Reply. Cassandra cluster automation in AWS with Terraform; Kubernetes API review on Google Cloud.
    • insparx. Architecture and coding of a sensitive-data migration from MySQL to Postgres and MongoDB with Spring Batch.
Founder and CIO, powerdata GmbH 2020 Q2 — 2025 Q4

Led 15 engineers and 100+ AWS accounts; AI call-center automation handling ~50,000 calls a day. Acquired in 2024.

  • Berlin SaaS company for the energy sector, building serverless SaaS products on AWS for German power and utilities companies.
  • Built and maintained an omnichannel customer-support automation platform: voice, chat, and email intent recognition, with the routine use cases automated behind it. Deployed at one of the largest German power and utilities companies, it handled around 50,000 calls a day: meter readings, banking changes, payment adjustments, all resolved end to end without a human agent.
  • Led a team of 15 engineers and managed all AWS infrastructure, more than 100 accounts, following AWS security and compliance best practices.
  • Owned product and project architecture, infrastructure, and DevOps planning; every solution passed AWS architectural reviews. Spoke about the work at AWS Summit and an energy industry symposium. The company was acquired in 2024.
Founder and CTO, connectAI 2021 Q2 — 2023 Q2

Co-founded an omnichannel AI customer-care platform for the German energy industry.

  • Co-founded an omnichannel customer-care platform for the German energy industry. Voice, email, chat, and document recognition handled through one system.
  • Self-learning digital assistants on AWS that handled routine energy cases on their own: meter readings, tariff changes, relocations, banking and payment updates. Personalised and proactive, on the channel the customer chose.

Selected Projects

OpsWorker AI for ops
  • AI investigation layer for Kubernetes alerts. “AI that accelerates understanding, not replaces it.”
Beyond Chat: AI Workflows That Actually Investigate Alerts Watch the talk on YouTube
Call Center Automation 50K calls/day
  • AI-driven call center for one of the largest German power and utilities companies, handling around 50,000 calls a day end to end.
  • Conversations exposing the gap between how the tech industry presents itself and how it actually works.
Root Cause Podcast Watch on YouTube
  • Writing on automation vs. understanding, AI hype cycles, and engineering leadership.
TT Thoughtful Technologist Read the blog
  • icinga2-slack-notifications, an older monitoring-to-Slack integration I no longer maintain. It solved a concrete problem and found its audience in that niche (73 stars, 36 forks).

Skills

'09-11 '12-14 '15-16 '16-20 '20-25 now Architecture and serverless: Lambda, ECS, S3, DynamoDB, SQS, SNS, API Gateway, CloudWatch, Step Functions, Amazon Connect, and most core services; 100+ account org. Multi-department account design, management, and security; Azure Functions, Azure AI Foundry (formerly Azure OpenAI / AI Studio), Bot Framework chatbots, application gateways, Blob Storage. Multi-agent orchestration, MCP, Strands Agents SDK, Claude API, OpenAI API, Amazon Bedrock, AgentCore, RAG, memory management, LLM observability. CloudFormation, CDK, Terraform, Helm, Ansible, CI/CD pipelines, GitOps. Kubernetes, Helm, istio, observability, incident response, production operations. Amazon Connect, Twilio API, telephony and call-flow automation. Primary language today: agents, automation, and tooling. NestJS backends, Next.js/React frontends. Go services, CLIs, and performance tooling. Java, Scala, Groovy, Grails, Spring — the JVM years. Hadoop, HDFS, HBase, Presto, Storm, Elasticsearch, MongoDB, Cassandra. Engineering management, mentoring, hiring, 15-person team.

touched regular heavy daily work

Click a row for the concrete technologies behind it.

Cloud
AWS and Azure architecture, serverless architectures, GitOps. Azure: multi-department account design, management, and security; Azure Functions, Azure AI Foundry, Bot Framework chatbots, application gateways, Blob Storage
Infrastructure as Code
CloudFormation, CDK, Terraform, Helm
Operations
Production operations, Kubernetes, Helm, istio, observability, alert investigation, incident response, load testing
Contact Centers
Amazon Connect, Twilio API, telephony and call-flow automation
Languages
Python (primary), Go, TypeScript (NestJS, Next.js/React); earlier Java, Scala, Groovy
AI & Agentic
Multi-agent orchestration, MCP (Model Context Protocol), Strands Agents SDK, Claude API, OpenAI API, Amazon Bedrock, AgentCore, RAG, memory management, LLM observability and token cost tracking
Leadership
Engineering management, mentoring, hiring

Certifications

Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization, DeepLearning.AI 2024
  • Credential ID WS2YY5PASRJ0
Neural Networks and Deep Learning, DeepLearning.AI 2024
  • Credential ID VIV4QOZKJFBX
AWS Partner: Generative AI Essentials (Business), Amazon Web Services (AWS) 2023
Registered Yoga Teacher, Yoga Alliance 2023
AWS Certified Solutions Architect – Professional, Amazon Web Services (AWS) 2021
AWS Certified Solutions Architect – Associate, Amazon Web Services (AWS) 2016

Education

Master's degree, Mathematics, Yerevan State University 2010 — 2012
  • Graduated with honors, top grade on the 20-point scale.
Bachelor of Science, Mathematics, Yerevan State University 2006 — 2010

Languages

Speaking & Writing

  • Interviews with senior technologists on the decisions and truths behind the hype.
Root Cause Podcast Watch on YouTube
  • Essays on engineering, automation, and AI.
TT Thoughtful Technologist Read the blog
  • Beyond Chat: AI Workflows That Actually Investigate Alerts. Why chat-based AI is the wrong shape for incident investigation, and what works instead: workflows that investigate like a teammate who knows the system, shown in a live demo against a simulated 3am alert.
Beyond Chat: AI Workflows That Actually Investigate Alerts Watch the talk on YouTube
  • Agile in Tradition: how to iteratively build an automated call center solution handling 50,000 calls per day.
AWS Summit Berlin 2024 — Speaker 2024
  • AIM207: what’s next in contact centers with Amazon Connect and generative AI. Level 200, Intermediate.
  • Automation for Energy Operators: what actually gets automated in energy customer service, from the powerdata work.