Episode

S12 Bonus: Tobias "Tobi" Konitzer, Growthloop

Podcast
Code Story: Insights from Startup Tech Leaders
Published
Mar 26, 2026
Duration seconds
2237
Processing state
processed
Canonical source
https://codestory.co/podcast/bonus-tobias-tobi-konitzer-growthloop/
Audio
https://pdst.fm/e/pscrb.fm/rss/p/audio4.redcircle.com/episodes/b950858a-cf01-49da-aee9-93e503c65b77/stream.mp3
JSON
/v1/public/podcasts/code-story/episodes/s12-bonus-tobias-tobi-konitzer-growthloop
Markdown
/podcast/code-story/s12-bonus-tobias-tobi-konitzer-growthloop.md

Actions

  • POST https://stenobird.com/v1/public/podcasts/code-story/episodes/s12-bonus-tobias-tobi-konitzer-growthloop/transcription-requests
    Idempotently request low-priority transcript generation for this episode.
  • GET https://stenobird.com/podcast/code-story/s12-bonus-tobias-tobi-konitzer-growthloop.md
    Read the agent-friendly Markdown representation of this episode resource.

Summary

Tobi Konitzer shares how he transitioned from building complex Bayesian models in a vacuum to creating an autonomous decisioning system at Growthloop. He explores the necessity of aligning technical vision with customer feedback and the shift from marketing tools to agentic AI.

Topics

  • Agentic AI
  • Reinforcement Learning
  • Causal Inference
  • Startup Strategy
  • Marketing Technology
  • Product Management
  • Decisioning Systems
  • Bayesian Modeling

Highlights

  • Failure mode: Building highly sophisticated models in isolation without validating market demand or customer needs
  • Practical takeaway: Use a 'product growth map' to sequence technical requirements like causality data and agentic context graphs
  • Main idea: The future of marketing lies in moving from descriptive machine learning to autonomous, outcome-optimized decisioning networks
  • Strategic insight: Avoid over-reliance on model accuracy; focus instead on end-to-end pipelines that generate measurable returns
  • Leadership lesson: For highly specialized roles, avoid recruiters and focus on finding talent capable of handling complex, non-commoditized problems

Chapters

  1. 1:00 The Trap of Technical Perfection: Tobi reflects on the failure of building complex Bayesian reinforcement learning models that ultimately had no market demand.
  2. 7:30 From Tooling to Autonomous Decisioning: The vision for Growthloop: shifting from simple marketing tools to an opinionated, outcome-optimized decisioning network.
  3. 10:50 Building with Customer Feedback: How to avoid building in a vacuum by using existing primitives and integrating customer feedback into the product roadmap.
  4. 20:40 Validating the Agentic Vision: The process of aligning high-level AI vision with market reality through structured communication and documentation.
  5. 24:10 The Architecture of Agentic AI: A deep dive into causality data, agentic context graphs, and the infrastructure needed for automated traffic allocation.
  6. 34:20 Lessons in Machine Learning Utility: Why accuracy is the wrong metric and why end-to-end pipelines for measurable returns are what actually matter.
  7. 37:50 The Future of Agentic Commerce: Defining a world where AI agents govern the entire checkout and commerce process.