Resume

Daniel Yakobian

Senior software engineer in Los Angeles. Labeling and ML data platforms at Snap, payments and search at Clutter, trust and safety tools at Tinder.

Download PDF 1 page · 29 KB

$47Ma year in avoided labeling cost at Snap

85%lower cost per label with 4 outside vendors

15 to 2weeks of labeling turnaround

60%fewer bots at Tinder in 3 months

Experience

Apr 2020 –
Jan 2025

Snap Inc. · Los Angeles, CA

Promoted from L3 to L5 (Senior); tech lead for a team of 5 engineers, presenting plans and OKRs to leadership.
Led a labeling platform that brought on 4 outside vendors, cutting cost per label 85% (about $47M a year in avoided cost) and turnaround from 15 weeks to 2.
Raised label quality 40% (labeler agreement, Krippendorff’s alpha) by leading SWE, SRE and vendor teams and building dispute-resolution tools.
Implemented ML-based NSFW detection, reducing complaints by 20% (TensorFlow, PyTorch, OpenCV).
Optimized ad scene-break detection in user videos, integrating with the Ads APIs using low-level FFmpeg.
Managed Kubernetes deployments and Prometheus/Grafana alerts; wrote playbooks; primary on-call.
Mentored junior engineers and led biweekly data-engineering knowledge shares.

Jul 2018 –
Mar 2020

Clutter Inc. · Los Angeles, CA

Refactored payment infrastructure and built new checkout, improving conversion rates and payment reliability.
Developed storage search engine backend/frontend (Node.js, Spark, DynamoDB, Lambda, React).
Rebuilt customer-facing pages (clutter.com) using Ruby on Rails, React, TypeScript, GraphQL.

Jun 2016 –
Mar 2018

Tinder · Los Angeles, CA

Built the “download my data” portal from scratch, cutting account deletions by 4% (React, Node.js, DynamoDB).
Reduced bot presence by 60% in 3 months through data analysis (Node.js, Kafka, DynamoDB, Spark).
Built internal tools for bot, photo and user management (React, Node.js, TypeScript).

May 2015 –
May 2016

Dyad Medical Inc. · Los Angeles, CA

Built cloud-based medical imaging analysis tool integrating C++ analysis app with Node.js via RabbitMQ.

Nov 2013 –
Dec 2015

Colign · Los Angeles, CA

Scraped & indexed 30GB of patent data, built search engine with custom XML parser & Elasticsearch.

Languages: Go, Python, TypeScript/JavaScript, Java, Scala, SQL, Ruby
Frameworks: Node.js, React, Ruby on Rails, GraphQL
Cloud / tools: GCP, AWS, Kubernetes, Docker, Airflow, Dataflow, Spark, Prometheus, Grafana
Databases: Spanner, Bigtable, DynamoDB, MongoDB, Redis, Elasticsearch
Streaming: Kafka, RabbitMQ, SQS

May 2016

Case Western Reserve University · Cleveland, OH