Skip to content
View kgkazakos's full-sized avatar

Block or report kgkazakos

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
kgkazakos/README.md

Kostas Kazakos

Computational Product Researcher · PhD in HCI · Melbourne, AU

I build AI-native research infrastructure — systems that keep humans in command while extending the reach of qualitative and quantitative insight.

Computational Product Research (CPR)

CPR is the bridge from static insight to agentic intelligence. I architect open-source tools using Python, LangChain, DSPy, and LLMs that make research rigorous, transparent, and scalable.

Project What it does Status
CausalTrack Multimodal Say-Do Gap detection · Gemini + Neo4j · 611-node knowledge graph CHI 2027
Synthetic User Council Agent-led ecosystem simulation for pre-deployment product research CHI 2027
transcript-cleaner Fix messy interview transcripts · rule-based + model-agnostic LLM Live

40 orbital tools shipping every Friday through Dec 2026 · full roadmap →

Stack

Python LangChain DSPy LangGraph Neo4j Streamlit Gemini Claude OpenAI pytest GitHub Actions

Writing

Currently writing The AI-Native Product Researcher: Scaling Empathy and Rigor — a book on the transition from episodic insights to continuous intelligence.

Lead Researcher · ex-Google L6 (Transformative Impact) · Atlassian · PhD HCI University of Melbourne

Pinned Loading

  1. causaltrack causaltrack Public

    Audio-based behavioral truth detection for user research - A methodology validation study

  2. kgkazakos.github.io kgkazakos.github.io Public

    Kostas' Personal Github Site

    HTML

  3. synthetic-user-council synthetic-user-council Public

    A multi-agent AI system that challenges researcher assumptions before user studies begin.

  4. transcript-cleaner transcript-cleaner Public

    Fix messy interview transcripts in seconds — speaker labels, filler words, fragmented turns, and broken timestamps. Rule-based by default. LLM-powered when you need it.

    Python

  5. notebook_lm_prompt_generator notebook_lm_prompt_generator Public

    A command-line Python tool that acts as an AI creative director for Google's NotebookLM. It automatically checks your ideas against prompt engineering best practices and generates highly optimized …

    Python