Contact

Egill Erlendsson

Senior Software Engineer · AI automation builder · Technical lead

Your Partner for Automated AI Engineering.

I put the AI in Get Things Done.

Stylized portrait of Egill Erlendsson
Fig. 01 ai-gill.png · rendered 2026

Profile

Senior software engineer and technical lead with 20+ years of experience building, modernizing and operating production-critical systems across backend services, APIs, data platforms, real-time workflows, cloud infrastructure, hardware integrations and business automation.

Since September 2025, almost all of my hands-on project work has been done with AI/LLM tooling as a core part of the engineering process. I started with IDE-assisted workflows, but increasingly moved into console-based AI workflows: Claude Code, Claude SDK, Codex CLI, custom scripts, task runners, GitHub-task-driven automation, iterative improvement harnesses, test generation, codebase analysis, API/client generation and workflow-specific tooling.

My strongest pattern is using AI to move faster without losing engineering control: breaking down messy operational problems, building prototypes quickly, grounding the work in tests/logs/telemetry, and turning the useful parts into reliable production systems. I am especially interested in AI agents and automations that do real work inside companies, while respecting data protection, access control, observability and operational safety.

AI / LLM engineering & automation

Experience

Lead Developer — RSF Iceland

Present

Built and evolved a business-critical digital platform for Icelandic fish markets, covering the full operational lifecycle from supply registration and live auctions to transactions, settlements, invoicing, payments, reporting and operational administration.

The platform supports national fish auction workflows and is used by buyers, suppliers, market operators, finance teams, government institutions/agencies and administrators in a domain where correctness, reliability and operational continuity are critical.

Selected work

  • Designed, maintained and evolved backend systems supporting live auctions, financial workflows, invoicing, settlements, reporting and operational administration.
  • Led major modernization work across the platform, including migration from Rails 4 / MS SQL Server to Rails 7 / PostgreSQL on AWS.
  • Designed and built production APIs consumed by customers, government institutions/agencies, mobile applications and internal operational tools.
  • Worked across Rails, Python/FastAPI, Elixir/Phoenix, .NET Core, Redis, PostgreSQL, MS SQL Server, AWS, Linux, Docker, Terraform and CI/CD.
  • Built and maintained Python/FastAPI services for API gateway and backend integration workloads.
  • Used Python, Polars, Pandas, PostgreSQL and MS SQL Server to process, reconcile and validate operational and financial datasets.
  • Diagnosed and resolved data correctness, rounding, performance, reliability and integration issues across production workflows.
  • Used AI/LLM tooling for codebase analysis, implementation planning, refactoring support, test generation, debugging and structured development workflows.

Relevant technologies Ruby on Rails, Python, FastAPI, Elixir/Phoenix, C#/.NET Core, PostgreSQL, MS SQL Server, Redis, AWS, EC2, RDS, Docker, Linux, Terraform, CI/CD, Polars, Pandas, OpenAPI

Senior Software Engineer / Consultant — Independent Consulting

Present

Worked as an independent contractor on selected consulting engagements ranging from long-running product/platform projects to focused interventions around backend services, infrastructure, integrations, data, automation and production reliability.

Since September 2025, this work has been heavily AI-assisted, using Claude Code, Claude SDK, Codex CLI, Cursor, custom scripts and console-based workflows to accelerate implementation, testing, debugging, integration work and operational tooling.

Selected work

  • Avero — silent retail loss prevention. Solo-built a sensor-to-gate retail loss-prevention system across 40+ stores, tracking 1M+ customer movements per month — full pipeline from 3D ceiling sensors through edge processing (~100 events/sec) to silent gate decisions, signal-to-decision in milliseconds.
  • Migrated edge software from Go to Rust after AI-assisted analysis surfaced lock contention and runtime delays; identified and patched a memory leak in the production MQTT library along the way.
  • Built signal-recording and replay harnesses so AI-driven refactors and algorithm changes are validated against captured real-world traffic, with daily AI QA passes that score live behaviour (track stitching, group handling, tailgating detection) against expectation each release.
  • Built AI-managed release flow and incident triage — humans only get paged when AI judges a real call is needed, keeping the system silent until it actually shouldn't be.
  • Skvass — remote gym access automation. Solo-built a Raspberry Pi access bridge for a squash gym connecting a third-party booking platform with biometric scanners and physical gate hardware — designed, built and operated remotely from another country, never on-site.
  • Owned the architecture: canonical mapping between booking-platform identifiers and biometric enrolment IDs, changed-only sync to preserve device biometric state, force-all recovery, and QR / NFC / digital-wallet credential paths.
  • Used AI to read the booking-platform API and the scanner WebSocket protocol from sparse documentation and observed device messages, and as a remote hardware-debugging partner — a "scanner not opening the gate" investigation was traced through logs and command traces to a wiring issue diagnosed from a country away.
  • ComeUnity — Danish university community platform. Built and evolved a Laravel 12 / Statamic 6 backend / CMS powering a Danish university community mobile app and company portal.
  • Built substantial parts of an Expo / React Native mobile app, including API integration, authentication / profile flows, messaging / social features, content delivery and deep linking.
  • Designed and built a semi-automated AI-assisted OpenAPI client workflow that keeps the mobile app's API layer aligned with the published contract — generated client, hand-written wrappers, TypeScript types, mocks and tests all move together each time the backend changes, with AI inspecting each contract diff for anomalies and Figma context linked through Claude Code so design-driven backend changes land with their UI implications already visible.
  • Quake Monitor — earthquake exhibition at Harpa Concert Hall. Built the live earthquake visualisation now running on an 85-inch 4K screen inside the volcano exhibition at Harpa — custom Iceland map, real-time projection, exhibition admin and AI-built calibration tooling. Shipped end-to-end without ever visiting the venue, with installer photos and AI-assisted QA standing in for site visits.
  • Built AI-assisted tools and custom-code workflows for sigrun.com — plugin / custom-code support and website maintenance / debugging automation.
  • Worked on a Swiss public transportation platform integrating backend software with real-time passenger information systems, signs / screens, routes, delays, alerts and operational messaging.
  • Acted as CTO / senior contractor for GetLocal, leading development of a SaaS e-commerce and site builder platform.
  • Worked as Tech Lead / consultant for the City of Reykjavík, helping shape in-house software development and delivery of municipal digital services, including a citywide one-click school enrollment solution.

Relevant technologies Python, FastAPI, Rust, Go, Ruby on Rails, Laravel, Statamic, React Native, Expo, TypeScript, OpenAPI, PostgreSQL, TimescaleDB, SQLite, Redis, Prometheus, Grafana, MQTT, WebSockets, Docker, Linux, nginx, AWS, Azure, CI/CD, Raspberry Pi, TCP/IP, RS485, Wiegand

Senior Software Engineer — Automattic

Worked in high-scale production engineering on a custom ad network and related analytics/reporting systems.

Selected work

  • Wrote and maintained code serving ads across a very large production network.
  • Worked on vendor integrations, analytics/reporting tooling, backend systems and operationally significant production workflows.
  • Contributed to systems where performance, reliability and scale were central concerns.
  • Worked in a distributed, asynchronous, remote-first engineering environment.

Relevant technologies PHP, JavaScript, backend systems, analytics/reporting, vendor integrations, high-scale production systems, remote/asynchronous collaboration

Founder / CTO — Grapewire

Built streaming, e-commerce and digital delivery systems for music and media, later pivoting into a Spotify-like service integrated with a major telecom company in Iceland.

Selected work

  • Built media streaming and digital delivery platforms.
  • Designed backend systems for catalog management, purchases, streaming and telecom integration.
  • Worked with major music labels, artists and commercial partners.
  • Built infrastructure-heavy systems involving encoding, media delivery, ticketing and digital commerce.

Earlier roles

  • Senior Software Engineer — Industria. IPTV solutions, mobile streaming over 3G, user-generated content portals and large-scale encoding in an infrastructure-heavy telecom environment.
  • Senior Software Engineer — National Bank of Iceland. C#, .NET and Oracle in securities, treasury and foreign-exchange related systems.
  • Senior Software Engineer — Kaupthing Bank. Internal and external web systems mainly in C#, .NET and MS SQL Server.

Technical skills

AI / LLM tooling
Claude Code, Claude SDK, Codex CLI, Cursor, GitHub Copilot, AI-assisted development workflows, console-based automation, GitHub-task-driven workflows, iterative code-improvement harnesses, prompt/workflow design, agentic automation, replay/analysis harnesses, RAG concepts, local/open model evaluation
Backend / server-side
Python, FastAPI, Ruby on Rails, C#/.NET, .NET Core, Rust, Node.js, PHP, Laravel, Elixir/Phoenix
APIs / integration
REST APIs, OpenAPI, service-to-service integration, API gateways, backend integration, MQTT, Socket.IO, TCP/IP, vendor integrations
Data / databases
PostgreSQL, TimescaleDB, MS SQL Server, Redis, Oracle, MySQL, SQLite, MongoDB, relational data modelling, operational data systems, time-series data
Cloud / infrastructure
AWS, EC2, RDS, S3, Azure, Docker, Linux/Ubuntu, Terraform, CI/CD, Ansible, nginx, systemd
Reliability / observability
Prometheus, Grafana, production troubleshooting, performance diagnosis, deployment hardening, monitoring, logging, operational support
Frontend / product
React, Next.js, React Native, Expo, JavaScript, TypeScript
Device / operational systems
IoT devices, sensors, physical gates, Raspberry Pi, RS485, Wiegand, industrial printers, VPCs, VPNs, DMZs, routing, firewalls

Languages

  • Icelandic — Native
  • English — Fluent / near-native
  • Danish — Fluent
  • German — Basic

Education

University of Reykjavík
Computer science / related technical studies · 2012 – 2014
Degree not completed

Portfolio

Selected case studies — production builds where AI is part of how the work actually gets done.