Kerem Gokhan

Booking.com B.V. — Software Engineer, Attractions

AmsterdamDec 2018 – Present

Attractions Search (2025 – Present) — backend on OpenSearch

• As part of a 4-engineer pilot team experimenting with AI-first software delivery practices, delivered Booking.com's Early Booker Deals campaign — designed the discount-configuration model, shipped the first prototype across multiple services, then drove the Search domain end-to-end including integration with a cross-sell platform owned by another engineering organization. Delivered ahead of schedule through broad production rollout, using experimentation to validate the change.
• Led the multi-year Attractions Perl deletion taskforce as its original strategy author. With the team, built and iteratively refined an AI-powered deletion workflow over 20+ iterations. Drove retirement of a large volume of legacy Perl code and scheduled jobs at organization-wide scale — clearing a major prerequisite for the cloud database migration. The strategy and tooling are now used by other engineering organizations and business units.
• Led the shift of the POI ↔ product mapping V2 launch from a risky cutover to a 3-phase non-inferiority experiment. Convened the supply data-science team and Geo partners to fix upstream data quality (triggered a Geo data-cleanup effort). Replaced a static mapping with a daily-refreshed model that substantially expanded coverage.
• Drove broad adoption of AI-assisted workflows. Created a one-step setup wizard that installs Claude Code with the Search API's Model Context Protocol (MCP) server and the organization's shared AI plugin out of the box; maintain that plugin repository as a library of core skills used across multiple business units and roles, and standardized AI-readable engineering practices across many team repositories.
• Built an internal desktop client for Claude Code adopted by designers, product managers, content managers, and engineers; coordinated testing, packaging, and contributions across multiple teams.
• Helped migrate geographic search from static city-to-product mappings to latitude/longitude geo search; built Multilingual Autocomplete to improve search recall and destination discovery; designed and implemented the backend for POI landing pages; shipped a substantial expansion of SEO landing-page coverage.
• Other Search work included extended availability-window improvements and Genius cross-sell scaling.
• Delivered the first Dropwizard → Spring Boot 4 / Java 25 migration in the engineering organization; now templated as a shared skill used by other teams for their migrations.

Attractions Catalog (2022 – Jan 2025)

• Designed and pitched a serverless machine-translation system from scratch. Secured the engineering director's buy-in, planned the rollout, and led 3–4 engineers to deliver on time. First production launch covered a major supplier catalog.
• Two years on, the system runs reliably across the full Attractions catalog and all supported content languages, and was extended to LLM-based English content generation for long-tail suppliers — unblocking many additional products, reducing dependency on manual content operations, and improving supplier-content scalability.
• Co-designed the Canonical Model (provider-agnostic raw content) and a serverless AWS pipeline (Lambda / CDK) that distributes raw content into Search, Product Catalog, and Distribution. Validated through experimentation before full rollout.
• Unblocked a previously stalled supplier-automation initiative; enabled large-scale automated supplier-catalog ingestion, turning a limited manual onboarding path into a scalable catalog-growth pipeline.
• Designed and built the visibility-history layer + automated product return-to-online flow: reduced manual recovery work and improved product return-to-online reliability.

Attractions Connectivity (Dec 2018 – 2022)

• Built and maintained 9 third-party API integrations with major travel-experience suppliers and connectivity partners, covering inventory import, availability/rates caching, reservation, and cancellation flows. Stack: Java, Dropwizard, Kubernetes, MySQL, Kafka, Redis, Hive/Spark, GitLab CI.

Booking.com B.V. — Team Lead & Software Engineer, Accommodations (Personalization)

AmsterdamAug 2015 – Dec 2018

• Led a cross-functional team of 5–10 across backend, frontend, mobile, and UX (2017 – 2018). Owned the hotel comparison experience and persuasion-messaging surfaces (urgency, scarcity, social proof) on Booking.com; ran A/B experimentation strategy.
• Hands-on people management: 1:1s, mentoring, performance reviews, hiring loop (programming & system-design interviews).
• Earlier years (2015 – 2017): graduate-program rotations across internal tooling, mobile API infrastructure (Google Now integration), and product experimentation. Stack: Perl, Python, Java, MySQL, Spark, Hadoop, Redis.

Earlier roles

• Comodo (Ankara) — Software Engineer, Browser team. Android and Chromium-based browser development. Oct 2014 – Jul 2015.
• Semantica Inc. (San Francisco, remote) — Frontend Developer. Favogue, a fashion social network (AngularJS, Ruby on Rails). Jun – Sep 2014.
• GamaPrime Studio (Ankara) — Product Designer & Android Developer. Augmented-reality treasure-hunt game on Android. Oct 2013 – Jun 2014.