Tomass Lacis · Independent Systems & Platform Architect
Complex systems.
Clear solutions.
I help companies put AI to work in their engineering, and architect, modernize and scale the backend platforms, distributed systems and cloud infrastructure underneath it.
AI Adoption • Distributed Systems • Cloud Platforms
- 12+
- years building production systems
- 1–5M
- devices & sensors monitored on platforms I led
- 10k+
- smart-energy hubs in an IoT fleet I built
- 3–5/day
- releases, up from 1–2 per week
AWS Certified Solutions Architect – Professional · MSc Computer Science
Problems I solve
Make better technical decisions.
Build systems that last.
AI is a demo, not a workflow
Your team experiments with AI, but it never reaches the repos, tickets and operations where the work actually happens.
The platform is fighting you
Deploys are slow and scary, environments drift, and every change needs a hero.
Costs grow faster than usage
Cloud bills and operational overhead climb while nobody can say which part of the architecture is to blame.
Scale is starting to hurt
Telemetry, devices or data volumes outgrew the MVP design, and reliability is slipping.
Big decisions, no architect
You need senior judgment on a design, vendor or migration, without hiring a full-time architect.
Hardware meets cloud
Device identity, provisioning and fleet management need to be production-grade, not a shared API key.
Services
How I can help
AI Adoption & Agentic Engineering
Put AI to work on real engineering and operations, not just demos.
02Architecture & Technical Leadership
Senior judgment on the decisions that are expensive to get wrong.
03Cloud & Platform Modernization
Simpler, cheaper, safer delivery from commit to production.
04Backend, IoT & Data Platforms
Telemetry, devices and data pipelines that hold up at scale.
- AWS
- Kubernetes
- Terraform
- Python
- Go
- Kafka
- TimescaleDB
- Claude Code
- MCP
Selected work
Proof, not promises
Own practice · AI engineering · 2025 – present
Running an AI engineering team of agents
A self-hosted team of coordinated AI coding agents that manages repos, docs, tasks, architecture, implementation, testing and conventions across a real product.
- Multi-repo
- repos owned by dedicated agents
- 24/7
- persistent agent sessions
Professional experience · Connected networking · 2026 – present
The cloud control plane for a wireless networking product
Zero-to-one cloud platform for a hardware startup's base stations and relay nodes, from AWS foundations and device identity to telemetry, OTA updates and an AI copilot for network engineers.
- 0 → 1
- cloud platform owned end to end
- AI + MCP
- chat-driven, human-approved device configuration
Professional experience · Telecom · 2016 – 2023
Telecom network monitoring at massive scale
Lead engineer on a monitoring platform ingesting 20–30 million data points every few minutes from millions of network devices and sensors, across on-prem data centres and AWS.
- 20–30M
- data points every few minutes
- 2,000+
- platform users
How I work
Assess → Design → Build → Measure
- 01
Assess
Understand the system, the team and the constraints. Find the expensive problems first.
- 02
Design
Make the key decisions explicit, with trade-offs, costs and a realistic roadmap.
- 03
Build or guide
Implement the hard parts hands-on, or steer your team through them.
- 04
Measure
Agree on what better looks like, then prove it with numbers.
Engagements are either a scoped project with clear deliverables, or fractional advisory a few days a month. A good first step is a fixed-scope Architecture & Platform Health Check.
Have a difficult technical problem?
Tell me what you're building, what's not working, or which decisions you're trying to make. I usually reply within two working days.