Tech Lead across Cost Engineering (2 devs) and Reliability/Metrics at Snap Inc. Built platforms saving $100M+ in infra costs; drove initiatives across 18+ teams and 450+ services.

🚀 SnapConfig — Config Delivery & Rollouts

• Led Q1 design of SnapConfig's rollout system — cross-platform orchestrator integrating service mesh, Envoy, health, and K8s pod-bucket assignment. 450+ services, 13k+ rollouts/day.

• Designed partitioned rollouts: staged canary → primary → global pipelines with per-partition config overrides.

• Shipped automated rollbacks via service-mesh blue/green analysis; onboarded all 22 Tier-0 services with platform and service-mesh teams.

💸 Cost Engineering

• Led platform cost chargeback across 18 teams (180+ rules) attributing 30% of Snap's infra cost to consumer teams.

• Led 4 contractors to re-architect opportunity detection (serial DAG → parallel config-driven) — $9M/year in new savings.

• Chair cost-strategy syncs with Finance/GCP/AWS on commitments and vendor alignment.

• Designed automated cost-remediation identifying $250k/day — higher coverage than GCP/AWS tools.

• Architected hourly AWS EC2 commitment model ($40M/year); shipped 50+ workflows ($50M/year).

📊 Reliability & Metrics

• Unblocked Envoy metrics migration (30% of Snap metrics) from Graphite to Prometheus.

• Engineered cardinality tooling across 300M metrics / 90k monitoring rules; pinpointed top 15 teams (180M cardinality) for cleanup.

• Cut 30M cardinality and PubSub volume via cleanup initiative.

🛠 Developer Experience

• Event Tracking System correlating 80k+ daily events with service health — key to resolving 30+ Sev0/Sev1 incidents.

• Modernized Change Management replacing Google Docs (1,400 CMs, 30k+ views since launch).

👥 Senior-hiring interviewer; onboarded 6 engineers across cost/reliability in 2 years.

🔮 2026: GenAI agents (on-call, Slack, Change Management, Cost); metric pre-aggregation removing pod-id labels from Envoy/Prometheus (82% projected cost reduction).

WebXT Platform Core Team

(tl;dr: ask me about localization pipeline, or test automation)

• Part of the Platform Core team, working on infrastructure both on server side and client side to enable devs in the org to achieve more

• Primary contact for high business impact pieces: localization pipeline, fallback config resolver, market detection, server-side test and deployment automation --> +450 engineers and +25 canvases depend on these

• Backup for other high impact areas: App Diagnostics, Web Worker and more

Recent High Impact Projects:

• Built a Test automation pipeline to validate Server-Side PPE version against Client-Side Release Candidate version to ensure safer deployments

• Wrote a tool to reorder configuration priority of 30k files, which also automatically generated 50k validation scenarios for the newly generated 2000 files

• Redesigned Market Detection architecture to support two separate locales: display (~250 languages) and content (~60 regions)

• Added Staging step to Server-Side services

Previous High Impact Projects:

• Refactored to decouple the architecture libraries from react components to ease the transition to web components across the canvases

• Designed and implemented the localization pipeline from the moment the devs check in resx files to the moment the page consumes localized json config files

• Unified separate market detection logic that was used across WebXT pages: improving flighting and flexibility of the code

Even before:

• Contributed to performance optimization of the web page

• Designed and implemented adding inline scripts to the page dynamically without deployment

• Worked on design and migration of configuration files from CMS to Git

• Currently own localization pipeline which handles over 60 languages and 13k files

• Also designed and implemented a fallback config fetching/resolving library used across +25 different canvases

MSNews Service UI Team

• 4 months in MS News Service UI team building Launcher Mobile experience for Microsoft Launcher (+ Microsoft Duo) --> Moved to WebXT Platform Core team afterwards (handpicked by Dev Manager)

• Worked in a team of engineers to build a new Order Execution Management System(OEMS)

• Built one of the main components, dynamic UI generation based on FIX ATDL XML Schema

> Created two versions: C++ with MFC and Typescript with Vue+Vuex framework

> Processed files as large as 100,000 lines efficiently by caching a map of nodes

> Separated display creation from file reading, handling different type of XML Schemas

> Replaced XML inputs with pre-generated display objects for performance increase

• Co-operated on creating a C++ midlayer and Typescript with Vue+Vuex frontend to handle messages from Java backend

> Implemented Protobuf serialization and deserialization in midlayer and frontend

> Handled deserialized messages for replace, add, remove in a grid up to 1-million rows

> Added sorting, grouping and filtering functionality up to three columns for grids

• Participated in a proof of concept to prove frontend can handle the load of 1-million messages a second

• Experienced a switch to agile development as one of the first two teams in FactSet

• Trained with CMS systems, Drupal and Umbraco, implementing an HTML website into Umbraco

• Developed a C\# plug-in for form/chart generation from quiz results: implemented Azure SQL

• Migrated a web application into ASP.NET Core MVC and added multiple language support

• Led & Organized team meetings for a newer version of a Contracts Portal application