Members-Only
Recent Talks & Demos are for members only
You must be an AI Tinkerers active member to view these talks and demos.
June 17, 2026
·
Berlin
AI Voice Tutor for iOS
Learn German with an AI tutor that uses unique illustrations, real-time voice scenarios, and personalized flashcards generated from your spoken mistakes.
Overview
A German learning app for iOS. Every card is a unique AI-generated illustration, and instead of swiping through cards you can practice by speaking with a realtime voice tutor that drops you into dynamically generated scenarios – order a coffee, miss a train, argue with your landlord. It mines your conversation transcripts for the mistakes you actually made and turns them into new cards overnight. FSRS spaced repetition underneath, and I use it every day for my own German.
Tech stack
- SwiftUISwiftUI is Apple’s declarative framework for building native user interfaces across all Apple platforms with a single, unified codebase.This is Apple’s modern UI toolkit, leveraging the power of Swift to deliver a declarative syntax: you state what your UI should do, not how. It’s a unified system, meaning you write one set of code to deploy on iOS, iPadOS, macOS, watchOS, and tvOS (plus visionOS). Development is streamlined via deep integration with Xcode, offering features like real-time previews and automatic synchronization between the UI and app state. For existing projects, SwiftUI offers seamless interoperability with older frameworks like UIKit and AppKit, enabling incremental adoption without a full rewrite.
- GPT-Realtime 2GPT-Realtime-2 is OpenAI's advanced speech-to-speech model featuring GPT-5-class reasoning for low-latency, highly intelligent voice interactions.OpenAI's GPT-Realtime-2 elevates conversational AI by merging speech-to-speech capabilities with advanced reasoning. Designed for developers building complex voice agents, the model handles intricate multi-turn conversations, follows instructions with high precision, and executes parallel tool calling reliably. Key features include a 128K context window, configurable reasoning effort, and "preambles" (dynamic filler speech that eliminates awkward silence while the model processes complex tasks). It serves as a highly responsive, low-latency engine for real-time voice-to-action workflows.
- GPT Image 2OpenAI's state-of-the-art native image generation model delivering 99% text rendering accuracy and precise multi-object spatial reasoning.GPT Image 2 is OpenAI's latest leap in native image generation, engineered to solve the industry's toughest visual challenges. The model achieves a staggering 99% accuracy in text rendering (handling complex typography and multilingual labels natively) while rendering dense, multi-section layouts like infographics and UI mockups in a single generation. With native support for flexible aspect ratios up to 3:1 and high-fidelity 2K resolution, it delivers photorealistic outputs with accurate lighting physics and consistent character editing. It is the new benchmark for professional design pipelines, running twice as fast as its predecessor to turn complex creative briefs into production-ready assets.
- FSRSFSRS (Free Spaced Repetition Scheduler) is an open-source, machine-learning-driven algorithm that optimizes flashcard review intervals for maximum long-term retention.FSRS: Free Spaced Repetition Scheduler, is a modern, data-driven algorithm developed by Jarrett Ye to replace the legacy SM-2 method in spaced repetition software (SRS). It leverages a memory model—Difficulty, Stability, and Probability of Recall (DSR)—to precisely predict a user’s retention rate for each flashcard: this allows for highly personalized and efficient scheduling. Integrated into Anki since version 23.10, FSRS dynamically calculates optimal review intervals, often resulting in a significant reduction in daily reviews while maintaining a target recall rate (e.g., 90%). The algorithm is open source, ensuring transparency and continuous community-driven improvement.
- CloudflareCloudflare is the global cloud platform that secures, accelerates, and optimizes every internet connection: your network, applications, and teams.Cloudflare is an essential internet infrastructure and security provider. We operate one of the world's largest networks, proxying approximately 20% of the web and handling an average of 45 million HTTP requests per second. Our platform delivers integrated security (DDoS mitigation, WAF, Zero Trust) and performance (CDN, edge computing via Workers) from 330+ cities worldwide. This architecture ensures millions of internet properties—from small blogs to Fortune 1000 companies—remain fast, reliable, and protected against sophisticated attacks.