Lef Filippakis
@leffilippakis
I build paved-path platforms and agentic AI infra, making complex things boring in production and fast for product teams. Stockholm, remote-only.
What I'm looking for
I keep ending up building the same thing: the developer-experience layer on top of a complex platform, so other engineers ship safely and fast. Four roles, same pattern.
At Filed today (AI tax prep for CPA firms), I'm Principal engineer across infrastructure, AI platform, and backend. Three pieces of work I'd point to first: an agentic production-triage service tied behind a Slack-bot interface that consolidates logs and databases into a conversational DevEx (sandboxed, read-only, network-policied); reusable GitHub Actions workflows for LLM-powered PR review adopted across the org; and a multi-region Gemini load balancer that moved productionrate-limiting from workflow-blocking to non-event.
Before that at Visa, I built Ratchet: an internal MLOps framework that bundled IaC, a CLI, and CI/CD into a single Python tool. Data Scientists shipped regulated-industry ML to GCP Vertex AI through a transparent framework, maintaining a YAML config while GitHub Actions handled the rest. Service standards, deploy-and-rollback discipline, and observability defaults baked in. I gave two public talks on the design at NDSML Summit and the Stockholm MLOps Community meetup.
Earlier at Northvolt, I built the company's most-used internal Python library: a multi-backend data-access SDK across Redshift, ElasticSearch, S3, PostgreSQL, Timescale, and Athena. Daily-driver for 30+ engineers; organic adoption beyond the original team with feature requests; production-line incident debugging on the factory floor. I grew the team from 1 to 5 and founded the internal data community (0 to 100+ engagedemployees per month).
What I'm looking for: fully-remote Senior, Staff, Lead, or Principal IC roles.
Strong-fit profiles: AI-infrastructure platforms (Modal, Grafana, Databricks-shape), developer-experience teams, agentic-systems work, observability craft. Python-first (advanced async, libraries with broad userbases), TypeScript, Go (working).
Outside the dayjob: MSc in Complex Adaptive Systems and Engineering Physics from Chalmers; NordSec 2018 publication on predicting real-world software exploits; Arcturus, a browser-based virtual analogsynthesizer built ~95% by Claude Code agents in an autonomous loop I designed; Tibberios, a home-energy management tool on GitHub.
Experience
Work history, roles, and key accomplishments
Led engineering across infrastructure, AI platform, and backend for an AI-powered tax preparation platform for CPA firms. Built a multi-cloud (Azure AKS + GCP Vertex AI) platform with Terraform, geo-redundant enterprise PostgreSQL, and SOC 2-ready controls, and shipped AI systems including a prompt-optimisation layer and an agent-based production triage pipeline.
Senior MLOps Engineer
Visa (via Tink acquisition)
Feb 2023 - Aug 2025 (2 years 6 months)
Built Ratchet, an internal MLOps framework packaging Infrastructure-as-Code, a CLI, and CI/CD to let data scientists deploy ML products to production at scale on GCP Vertex AI. Delivered ML engineering enablement using AWS + Vertex AI tooling (including Triton and Cloud Run) and presented at community and industry meetups.
Data Scientist
Northvolt AB
Mar 2021 - Feb 2023 (1 year 11 months)
Built a multi-backend Python data-access library (Redshift, Elasticsearch, S3, PostgreSQL, Timescale, Athena) used daily by 30+ engineers to support production incident debugging. Grew a team from 1 to 5, founded an internal data community (0 to 100+ engaged employees/month), and shipped a line-position prediction model that saved thousands of dollars per month.
Full-Stack Data Scientist
Rocker (formerly Bynk)
Sep 2019 - Mar 2021 (1 year 6 months)
Delivered interactive dashboards providing product insights to 3 Product Managers and Senior Management across Credit Card, Loan, and P2P offerings. Built and managed Python data pipelines to the BigQuery data warehouse ingesting from Google Analytics, Firebase, on-prem SQL Server, Braze, Segment, Kafka streams, and PostgreSQL.
Innovation Zone Leader
Cybercom Group (now Knowit)
Apr 2018 - Jul 2019 (1 year 3 months)
As Innovation Zone Leader, drove and supported in-house projects by interns and consultants, doubling the company’s machine-learning project count in under 3 months and producing 2 white papers. Led customer PoCs with organizations including SKF, WSP, Stena Metall, Volvo Cars, and Winningtemp, and trained 100+ employees and customers through presentations and workshops.
Education
Degrees, certifications, and relevant coursework
Chalmers University of Technology
Master of Science, Complex Adaptive Systems, Engineering Physics
2016 - 2018
MSc in Complex Adaptive Systems and Engineering Physics at Chalmers University of Technology (2016–2018).
Piraeus University of Applied Sciences
Bachelor of Science, Automation Engineering
2010 - 2014
BSc in Automation Engineering at Piraeus University of Applied Sciences (2010–2014).
Tech stack
Software and tools used professionally
Availability
Location
Authorized to work in
Website
lef.fyiSocial media
Job categories
Skills
Interested in hiring Lef?
You can contact Lef and 90k+ other talented remote workers on Himalayas.
Message LefFind your dream job
Sign up now and join over 100,000 remote workers who receive personalized job alerts, curated job matches, and more for free!
