gokalp - platform engineer

Software engineer at eGain, SF Bay Area.

Works across platform, backend, and full-stack: APIs, authentication, cloud services, and infrastructure.

Has also worked on frontends, blockchain integration, dashboards, and design tools, and has led a student team.

Writes notes while learning how things work.

• linkedin(1)
• github(1)
• x(1)

• eGainSoftware engineer · Jan 2026 - present

  An enterprise software company.

  • Designed and built an enterprise API gateway on Apache APISIX, with Lua, Terraform, and AWS.
  • Built backend services for routing, SSO/OIDC authentication, authorization, and rate limiting.
  • Added monitoring and operational dashboards with Prometheus and Grafana.

  conceptscloud services on AWS, backend development, API design, authentication and authorization, rate limiting, testing and load testing, infrastructure as code

  how I workedLearned AWS and a new stack on the job. Worked with product, QE, and cloud teams.

  tools: AWS, Apache APISIX, Lua, Python, React, TypeScript, Terraform, GitHub Actions

• PoolTextIntern · Aug - Sep 2024

  A platform that connects journal editors and authors and uses LLMs to help authors improve their papers.

  • An advanced web editor with TipTap and Mantine, with state in Redux Toolkit.
  • A .NET microservice that made the platform more modular and reliable.
  • Improvements to the editor's user experience.

  conceptsrich-text editing, LLM-assisted writing, a scalable service next to the frontend

  how I workedShipped features to real users. Worked with the team on requirements and architecture changes.

  tools: React, TypeScript, TipTap, Redux Toolkit, .NET

• MoralabsIntern · Jun - Jul 2024

  A mobile game and app development studio.

  • The frontend of a web project, with mock APIs for frontend testing.
  • A custom npm library that simplifies how clients call backend functions.
  • Cryptocurrency wallet integration with Ethers and MetaMask.

  conceptsreusable libraries, blockchain integration, component design

  tools: React, TypeScript, Mantine, Redux Toolkit, Ethers.js, MetaMask, npm

• TurboardIntern · Jul - Aug 2023

  An international business-intelligence software provider.

  • Dashboards for a large, frontend-heavy BI application.
  • Bug fixes and product improvements, with a team of engineers.

  conceptsdata visualization, turning customer feedback into product changes, component and props design

  how I workedTalked with customers to understand their needs and gather feedback. Worked closely with a team of engineers.

  tools: React, Vue, Apache ECharts, Vuetify

• SurDurTeam lead and full-stack developer. University capstone.

  Map apps recommend places. Drivers need recommendations along the driving route.

  • A full-stack Android app that recommends places along a driving route.
  • Led Agile development and Scrum planning for a cross-functional team.
  • Designed database schemas and secure REST APIs with JWT authentication.
  • Worked across the frontend and the backend. Deployed on AWS.

  stack: React Native (Expo), FastAPI, MySQL, AWS

• LLM Chat SearchPersonal project. Open source (MIT).

  Chat history with AI coding tools is spread over many places and hard to search.

  • A VS Code and Cursor extension that searches chats from several AI coding tools in one sidebar.
  • Fuzzy search, plus optional semantic search with embeddings (experimental).
  • A reader that shows the full conversation when you click a result.
  • Unit tests and CI with GitHub Actions.

  stack: TypeScript, VS Code extension API, Fuse.js

  code: repo

• Design StudioSolo project.

  Turn a text prompt into an editable UI design.

  • LLM layout generation, with a lot of prompt design and iteration, and JSON validation.
  • An infinite canvas with zoom and auto-layout.
  • A Figma plugin that imports layouts and maps design tokens.
  • Accounts, saved projects, and REST APIs.

  stack: Next.js, Express, MongoDB, Groq, Figma Plugin API, Tailwind CSS

  code: front, back, plugin

• ChessySolo project.

  A chess app with AI game assistance.

  • A full-stack chess app with accounts and AI assistance.
  • Tests and CI with GitHub Actions.

  stack: React, Express, MongoDB, Groq, chess.js, Tailwind CSS

  code: app, api

How an LLM request is served, one stage at a time.

1. rate limitstudied
2. tokenizenext
3. prefill
4. decode + cachestudied
5. sample

studied: KV cache, token rate limiting. next: tokens.

conceptscloud services on AWS; backend development; API design and gateways; authentication and authorization (SSO, OIDC); rate limiting and caching; testing, performance profiling, and load testing; infrastructure as code and CI/CD; monitoring and observability; prompt design and building reliable LLM features (structured output, validation); developer tools: indexing, fuzzy and semantic search; full-stack product development.

working styletalking to customers and turning feedback into changes; leading a small team with Agile and Scrum; learning a new stack quickly; working with product, QE, and cloud teams; shipping to real users.

tools: Python, TypeScript, JavaScript, SQL, Lua, React, Node.js, AWS, Terraform, Docker, GitHub Actions, Prometheus, Grafana

B.S. Computer Science (honor student), Bilkent University, 2025.