Kagan Demirhindi

Product experience platform that helps companies make product intelligence actionable at speed and scale.

• Core engineer on Pendo's AI agent, working end-to-end across multi-agent orchestration, skills, MCP tooling and eight workflows.

• Built Pendo’s internal agent evaluation tool from scratch, combining LLM-as-a-Judge, repeatable test suites, trace-based debugging and multi-backend comparison. It enabled engineers and PMs to make eval-driven decisions on model migrations, MCP tooling and launches.

• Led model strategy and regional deployment architecture for Pendo’s AI agent, delivering its model migration and expansion to Japan. Owned implementation and coordination with Legal and Platform Ops.

• Improved agent reliability through trace-led debugging, hallucination guardrails and tool/context fixes. Synthesised traces, logs, analytics and user feedback into recurring AI-assisted gap analyses that surfaced high-impact improvements.

• Accelerated Pendo’s AI transformation through hands-on delivery and onboarding across multiple teams. Named Pendo's AI MVP for Q2 2026.

The only project management platform built specifically for client work.

• Designed and built an AI onboarding flow that generated customised project plans, prioritised tasks and assigned work from user input, doubling onboarding completion rates.

• Built Teamwork’s AI Profitability Forecaster with TinyTimeMixer and Prophet, turning historical data into one-click revenue, cost and profit predictions on AWS.

• Ran full-stack growth experiments in Vue.js and Go, using LaunchDarkly for controlled rollouts and Pendo, HubSpot and Hotjar to measure impact.

London-based fintech startup. ePOS app that turns tablets and smartphones into powerful cash registers.

• Joined at the earliest stage and built the company’s first merchant onboarding integration, working directly with the partner bank from technical discussions through implementation.

• Shipped a React, TypeScript and Ionic POS app used by real customers, including custom Capacitor plugins that connected phones to the bank’s payment terminals.

• Built Go and gRPC microservices, containerised with Docker and deployed on Google Kubernetes Engine.

Software consultancy in Izmir, Turkey. Strong culture of learning and exploring new technologies.

• Built large-scale Scrapy and Splash pipelines that collected gigabytes of web data daily, then cleaned and prepared datasets in Python and Pandas for model training.

• Analysed large industrial datasets with Pandas, Elasticsearch and Kibana, turning raw operational data into actionable insights for researchers.

Sole author of “Makine Öğrenmesi ve Sağlık Alanındaki Çalışmalarda Kullanımı”, covering machine learning, generative AI and Python applications in health research.

“How to Train Your (Coding) Agent”: Explored the current state of LLM capabilities along with practical patterns for managing context windows, subagents, skills and agentic workflows (bash loops) for AI-first development.

“Fewer Juniors Today = Fewer Seniors Tomorrow”: Sparked widespread industry discussion on the long-term talent pipeline risks of AI hype in software engineering.

A two-phase gate locks design verdicts (right place, right abstraction, right thing to build) before line-level review begins. Separate evaluation bars, a finding-depth ladder and a grader pass keep the output high-signal, while stack-aware guidance and cross-repo reasoning catch issues beyond the diff.

Portable skills + hooks that give coding agents wall-clock time awareness (chronos), AI-native delivery estimates (timescale), productive disagreement (pushback), first-class git/PR workflows (shipit) and structured thinking (socratic). Multi-harness: Claude Code, Codex and others.

Graduated with a 3.11/4.0 GPA while working as a software developer from my second year. Through the Ege Entrepreneurship Society, organised company visits and campus events that connected students with industry. Learned early that building is the best way to learn.