Kagan Demirhindi

More than 14,000 companies use Pendo to create better software experiences, including Salesforce, Cisco, Verizon, HubSpot and Zendesk.

• Core engineer on Pendo's AI agent, the AI layer of a platform 14,000+ companies use. Took it from pre-launch to 2,000+ weekly and 10,000+ total users across multi-agent orchestration, skills, prompt and context engineering, tool calling, MCP tooling, knowledge retrieval and semantic / vector search. Public adopters include Datasite and Meevo.

• We kept changing the agent with no way to measure the effect, so I proposed and built Pendo’s evaluation tool (Python, React, PostgreSQL): LLM-as-a-Judge verdicts, 163 test cases across 12 suites, repeated runs with pass thresholds for non-determinism, and comparison across seven backends including MCP and LangGraph. Now the gate engineers, PMs and managers run before every model, MCP and launch decision.

• Took the product into Japan, from raising the opportunity to owning the model strategy and regional deployment architecture behind it: a platform-wide move from OpenAI to Gemini that made it possible under data residency rules, run end to end with Legal and Platform Ops.

• Cut error rate from 8.8% to 0.9% and issue rate from over 30% to 13% in three months, lifting weekly retention from 30% to 37%. Across 27,000+ conversations, reduced rage-prompt rate 61% and unsupported-request rate 24% through trace-led debugging, hallucination guardrails, feedback loops and knowledge-base retrieval (RAG) that took unanswered how-to questions to near zero.

• Drove AI adoption from my own team outward: we transitioned first, then I did the same for another office and contributed to most of Pendo’s company-wide AI initiatives, through hands-on delivery rather than training sessions. 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, turning logged time, cost and billable rates into one-click revenue, cost and profit projections with confidence ranges. Self-hosted the models on AWS and selected between TinyTimeMixer and Prophet based on how much history each site had.

• Ran full-stack growth experiments behind LaunchDarkly flags, measuring impact in Pendo and HubSpot and rolling winners out gradually.

• Built and shipped product across the stack in Vue.js, Tailwind CSS, Go and PostgreSQL, from user-facing features to the services behind them.

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

• Built Go and gRPC microservices containerised with Docker and deployed on Google Kubernetes Engine, and shipped the React and TypeScript POS app on top of them, including the native plugins that connected phones to the bank’s payment terminals.

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

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

• Built backend microservices and REST APIs in Java with PostgreSQL and GraphQL, containerised with Docker, and automated build and deployment pipelines.

• Built large-scale Scrapy and Splash pipelines collecting gigabytes of web data daily, preparing datasets in Python and Pandas for model training, and analysed industrial datasets with Elasticsearch for researchers.

Open-source skills for coding agents (Claude Code, Codex and others), each shipped with its own eval suite covering behavioural patterns, safety boundaries and baseline comparison.

senior-review settles architecture before any line-level review begins, cutting the nitpicks agents usually produce. agent-fix-loop drives trace-verified reliability iterations. chronos, pushback, shipit, socratic and timescale cover wall-clock time awareness, productive disagreement, git/PR workflows, structured thinking and AI-native delivery estimates.

Author of “Machine Learning and Its Use in Health Research”, a chapter 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.

Graduated with a 3.11/4.0 GPA while working as a software engineer from my second year alongside coursework. 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.