Roran B.
Software Engineer Lead
Experience
Sep 2024 - Present
1 year 3 monthsSoftware Engineer Lead
Shell
- Managed the development of a Rust microservice for Inventory and Order Management, covering over 2,700 oil & gas stations.
- Enhanced system reliability by introducing a delay-and-retry approach when connecting with 3PL providers, ensuring smooth operations during intermittent outages.
- Architected a robust AWS infrastructure using AWS CDK in Rust, incorporating services such as Lambda, EC2, S3, RDS, SNS, and SQS.
- Employed AWS EKS and ArgoCD for Kubernetes hosting of Docker Images.
- Utilized gRPC with Protobuf (grpc-ruby) for efficient inter-service communication.
- Seamlessly integrated DynamoDB and PostgreSQL for flexible data management.
- Streamlined the CI/CD pipeline using GitHub Actions, adhering to DevOps principles.
- Orchestrated Rust & API development (using actix-web or Rocket) with strong testing (using cargo test).
- Chose OAuth 2.0 (Doorkeeper) and enforced secure API design practices.
- Achieved >85% test coverage using Rust's built-in test framework and mockall for mocks.
- Implemented third-party payments (Stripe, PayPal) in the Billing microservice with official Rust SDKs.
- Transformed a PHP-based Social Management Service into Rust (using actix-web or Rocket), improving maintainability and developer velocity.
- Maintained the frontend (React 17 + Bootstrap).
- Implemented OpsGenie alerts and Datadog metrics, enabling quick and well-informed decision-making.
Oct 2023 - Oct 2024
1 year 1 monthSenior Software Engineer
Artisan AI
- Built a high-throughput Rust inference gateway (actix-web + grpcio + protobuf) on EKS/KServe/Triton with dynamic batching and request coalescing; p99 latency 210→98 ms, GPU utilization 42%→79%, sustaining 3.1k rps at <1% error rate; validated via Prometheus + Datadog APM.
- Designed RAG retrieval with pgvector (HNSW) + Redis and hybrid BM25+vector re-ranking; recall@10 0.62→0.83, MRR 0.41→0.58; online handoff deflection +18%; evaluated with held-out datasets and A/B flags.
- Implemented Rust feature & embedding pipelines (Kafka → Rust consumers (rust-rdkafka) → S3/Parquet) with idempotent writes and exactly-once semantics; freshness 24h→5m; peak backlog drain 3h→17m (Grafana).
- Orchestrated fine-tune/LoRA jobs with Argo Workflows + CUDA containers; added evaluation harness (win-rate vs. baseline, toxicity/PII checks, cost/request); best model win-rate +21%, cost −34%.
- Shipped guardrails and safety filters (prompt-injection heuristics, PII redaction, allow-lists, output caps) with OpenTelemetry correlation; unsafe output 0.9%→0.2%, false positives −37%.
- Rust performance deep-dives: perf, cargo flamegraph, tokio-tuner, YJIT tuning, connection pooling, semi-circuit-breakers, exponential backoff; hot-path CPU −28% and GC pause p99 −46%.
- Platform & DevOps: Helm + Argo CD progressive delivery (canary/blue-green), KEDA GPU autoscaling, Spot/flexible node groups, gp3 + S3 lifecycle tiers; infra spend −26% QoQ at 99.9% SLO; change-failure 3.2%→1.1%.
- Supply-chain & compliance: SBOM (syft/grype), Trivy, Cosign, Vault + IRSA; MTTR 44→14 min via runbooks, synthetic checks, burn-rate SLO alerts.
- Built a “prompt lab” console (React + Vue + gRPC-web) for dataset curation, prompt/version experiments, and online A/B rollouts; cycle time days→hours.
- Established model/feature registry & lineage (labels, datasets, versions, eval scores) with automated drift detection; prevented two regressions pre-prod and standardized rollbacks to <10 min.
Apr 2021 - Sep 2023
2 years 6 monthsDevOps Engineer
LetsGetChecked
- Replatformed core services from EC2/ASG to EKS (AWS) with Helm and cluster-autoscaler; cut deploy MTTR 32→11 min and change-failure rate 3.4%→1.1%, evidenced by release timelines and on-call reports.
- Standardized IaC with Terraform modules and Packer AMIs; integrated Vault for short-lived credentials and Consul service discovery; reduced config drift incidents 9→1 per quarter and sped environment spin-up from hours to <15 min.
- Built CI/CD for .NET and container workloads using AppVeyor + Helm chart promotion and policy-gated change control; implemented container/EC2 vulnerability management to meet compliance SLAs (critical CVEs closed ≤48h).
- Defined SLIs/SLOs across ordering, results, and eRx flows; tuned alerts with burn-rate policies; p99 on patient-ordering APIs improved 310→190 ms and actionable alert ratio rose 35%→78%, validated in the observability dashboards and incident reviews.
- Hardened Kubernetes to healthcare expectations (HIPAA/SOC-2 aligned): network policies, least-privilege IAM, ECR image scanning, secrets via Vault/KMS; supported CLIA-certified/CAP-accredited lab workflows handling PHI end-to-end.
- Cut infra cost ≈24% by right-sizing node groups, adopting mixed-instance/Spot for batch workloads, gp3 EBS migration, and S3 lifecycle tiers; savings verified in AWS Cost Explorer and monthly FinOps reports.
- Enabled EMR/FHIR and eRx integrations to operate reliably at scale by adding per-tenant Helm overlays, canary rollouts, and blue/green strategies; reduced integration rollback events from 7 to 2 per quarter and improved partner onboarding lead time by 40%.
- Authored runbooks, diagrams, and platform docs; led SRE workshops on Kubernetes, Terraform, Vault, and Consul, lifting team mean time-to-diagnosis from 45→18 min and onboarding time for new engineers from 6→3 weeks.
Aug 2015 - May 2021
5 years 10 monthsBackend Engineer
HubSpot
- Led a 50% backend scaling initiative to prepare for customer growth, ahead of the $117M SolarWinds acquisition.
- Migrated microservices to GKE and led internal Kubernetes enablement.
- Addressed critical gRPC issues in Kubernetes, optimized Rust services (refactors, memory & GC tuning), and increased stability during rolling updates.
- Used Diesel/SeaORM to streamline database access in Rust.
- Contributed to a robust MySQL platform serving >2,000 databases and tuned query performance.
- Leveraged MongoDB for high-volume CMS storage and retrieval.
- Contributed to reliability and feature expansion of Rust-based microservices in the CRM ecosystem.
- Built RESTful APIs in Rust (using actix-web or Rocket) and Node.js for seamless inter-service data exchange.
- Implemented Kafka consumption in Rust microservices using rust-rdkafka.
- Bridged SRE and Backend; maintained frontend with Angular 9; instrumented with New Relic.
Summary
Senior Rust Developer with over 10 years’ experience, strong DevOps skills, and deep knowledge of AI engineering.
I care about a healthy, collaborative culture—practicing clear communication, taking ownership, and fostering a blameless learning mindset—so teams move fast without burnout and deliver real value to people.
Languages
English
AdvancedIrish
AdvancedEducation
Oct 2013 - Jun 2014
Dublin City University
Master’s Degree · Computer Science · Dublin, Ireland
Need a freelancer? Find your match in seconds.
Try FRATCH GPT More actions
Similar Freelancers
Discover other experts with similar qualifications and experience
- English
- Deutsch
2025 © FRATCH.IO GmbH. All rights reserved.