Hi, my name is

Brave

I build the platform layer between cloud and product

Platform engineer at Weaviate Cloud. Multi-tenant PaaS, control planes, Kubernetes

About Me

I build cloud infrastructure platforms - the provisioning systems, lifecycle orchestration layers, and platform services that expose complex infrastructure as a reliable, scalable product.

Currently at Weaviate Cloud, owning platform capabilities end-to-end across a production multi-tenant PaaS on GCP, AWS, and Azure. My work spans the control plane, cluster lifecycle automation (Temporal), storage platform architecture, and GPU inference infrastructure - in a small, senior team where engineers own significant problems from design through production.

Previously at KPMG’s managed services practice, leading platform delivery for ISV clients in regulated industries: fraud detection, risk analytics, and compliance tooling deployed as managed services for banks, financial regulators, and government institutions.

CNCF Kubestronaut. Most effective in roles combining deep technical ownership with technical leadership.

My stack:
  • Go
  • Kubernetes
  • Temporal
  • gRPC
  • Pulumi
  • OpenTelemetry

Experience

Senior Platform Engineer - Weaviate
Oct 2024 - Present

Weaviate Cloud is a multi-tenant PaaS - managed vector database clusters provisioned on-demand across GCP, AWS, and Azure, serving billions of vectors.

  • Own and extend Go-based control plane services provisioning isolated database clusters per tenant across GCP, AWS, and Azure - covering VPC peering, Kubernetes bootstrapping, and cloud-specific networking and storage configuration
  • Designed and built cluster lifecycle orchestration using Temporal - full provisioning and teardown state machine with retry logic, timeout budgets, and rollback handling against partial cloud API failures
  • Built and optimise GPU-based embeddings service running inference models on GPU nodes - batching and request routing tuned for throughput and p99 latency under production load
  • Develop observability and profiling infrastructure with OpenTelemetry to monitor platform operations and database performance
  • Manage production Kubernetes clusters across three clouds, each with distinct CNI, IAM, and load balancer implementations
Cloud Architect - KPMG MBS
Apr 2022 - Sep 2024

KPMG managed services operated as an external platform provider - ISVs in fraud detection, risk analytics, and compliance tooling partnered with KPMG to run their software as a managed service for banks, financial regulators, and government institutions, with KPMG owning the full platform and operational layer. I owned end-to-end delivery of the platform layer for multiple ISVs - from initial architecture to handoff.

  • Built managed deployment platforms per engagement: custom Kubernetes controllers and helm charts, resource governance, multi-tenant isolation, and security policy across client environments
  • Designed platforms with least-privilege defaults, audit logging, and compliance controls satisfying FCA and SEC-adjacent requirements without bespoke work per client engagement
  • Built customer-facing portals and API surface through which clients monitored deployments, and used the services
  • Designed SLO/error budget framework across managed services - providing clients with contractually-backed service commitments and a structured mechanism for reliability trade-offs
  • Built observability stack (Prometheus, Grafana, Loki) and incident response automation that classified failure patterns and triggered remediation runbooks - reducing MTTR on recurring failure modes
Enterprise Architect, Cloud Infrastructure - eTranzact International
Oct 2021 - Apr 2022
  • Architected high-availability infrastructure achieving 99.9% uptime
  • Reduced system latency by 40% through architecture optimisation
  • Designed and implemented zero-data-loss disaster recovery systems
  • Implemented automated scaling and failover mechanisms
Cloud & Infrastructure Engineer - Various Clients
Jul 2019 - Oct 2021
  • Built GPU training infrastructure including a job scheduler balancing cluster utilisation against wait time without starving long-running jobs
  • Implemented checkpointing for training pipelines enabling mid-run recovery from failure, reducing wasted compute on long-running workloads
  • Built CI/CD pipelines with progressive rollouts, metric-gated promotion, and automatic rollback on regression
  • Built distributed tracing and monitoring for debugging multi-service distributed workflows

Projects

ConnectRPC AuthZ
go connectrpc grpc
ConnectRPC AuthZ
Authorisation interceptor library for ConnectRPC in Go. Supports custom authorisation logic and policy-based authorisation via Casbin, covering unary and streaming RPCs across Connect, gRPC, and gRPC-Web protocols.
Health Checker
go kubernetes docker
Health Checker
Lightweight Go utility for HTTP health checks in distroless container environments - built to solve a real operational gap in production Kubernetes workloads.

Get in Touch

Always happy to chat about platform engineering. Reach out at: