We strongly adhere to the Lean Startup and Agile principles, favoring action over analysis, imperfect iteration over perfect delivery, and face-to-face communication
As Senior Platform Engineer, you’ll be an individual contributor within the engineering team, occassionally embeded in squads, but most of the time working on the platform roadmap
You won’t lead people (yet), but you will lead technical initiatives that touch every engineer at haddock
You’re expected to be opinionated, pragmatic, and to contribute code
This is not an ops-only role
Haddock is an AI-first company, and so is this role
More than 50% of our infrastructure cost today is tied to AI workloads, not traditional SaaS software operations
You’ll own both worlds: classic platform work and the cost, performance and reliability of our AI stack
You’ll also use AI daily to do your own work, from infra automation, debugging, writing code, monitoring, etc
Secure our data layer (VPC, DNS, secrets, IAM)
Improve CI/CD and developer productivity
Make infra observable: clear, actionable metrics and dashboards, including AI workload cost and performance
Own cloud cost: visibility, predictability, optimization, with a sharp focus on our AI spend
Help us scale performance: PostgreSQL optimization, caches, queues, pub/sub, and search/analytics infra (Elasticsearch, BigQuery, ClickHouse…) when the product calls for it
Architect what comes next: alerting, team growth, next layer of scale
Moving our PostgreSQL and MongoDB databases behind a VPC with no downtime, and tightening secrets management and IAM along the way
Getting AI infra cost and performance under control: instrumenting our AI workflows, identifying waste, and making our AI spend predictable as we scale
Improving CI/CD and preview environments so product engineers (and AI agents doing the work) ship faster with less friction
Fixing our top performance bottlenecks: identifying the queries and endpoints that hurt most, fixing them, and mentoring the team so good query patterns become the default
Leading the decision on whether to adopt Elasticsearch, BigQuery, ClickHouse or similar, backed by data from real product needs