Megan

Social Sim Studio is a voice-first simulation tool that lets students practice real conversations before they happen. Most communication practice happens in controlled environments, but real life is unpredictable. Students don’t struggle because they don’t know what to say. They struggle because conversations don’t follow a script. This tool creates realistic, voice-driven scenarios where students can rehearse everyday interactions like ordering food, asking for help, or joining a conversation. The system responds in real time, asks for clarification when communication is unclear, and adapts tone based on personality and difficulty. Instead of static prompts, students interact with dynamic characters that behave like real people. I used ElevenLabs to power adaptive voice interactions. Each character’s voice changes based on friendliness, supportiveness, and pacing, allowing the same scenario to feel completely different depending on context. This creates a more realistic and flexible practice environment. Custom scenarios allow educators to tailor interactions to individual students, including setting, communication partner, and goal. The system then generates both the conversation logic and voice behavior dynamically. The result is a safe space to practice real-world communication in a way that actually prepares students for real situations. Built by a special education administrator for special education teachers and service providers to use with students who have challenges with communication and social interactions. Social Sim Studio provides a scaffolded way for students to practice conversations in simulated real life scenarios.

Submitted 9 Apr 2026

G8KEEPER is a real-time multiplayer social deduction game where you can’t reliably tell who is human and who is AI. Players are dropped into a shared environment with eight total players. Some are real players, others are AI-controlled agents that move, interact, and communicate just like humans. Each round, players explore rooms, complete objectives, and interact with fully voiced NPCs to gather information. The goal is to identify two things by the end of the game: your assigned room and the misaligned player (randomly assigned) who is actively trying to manipulate the outcome without being detected. What makes G8KEEPER different is how the AI behaves. Non-human players aren’t scripted, instead they simulate intent. They move with purpose, send messages, interact with NPCs, and occasionally mislead other players. This creates a constant sense of uncertainty where every decision feels like a risk. I used ElevenLabs to power all NPC and narrator voices, giving each character a distinct tone, personality, and presence. Voice is not just cosmetic here, it’s a critical part of how information is delivered, trusted, or questioned. On the backend, each game lobby is powered by a Cloudflare Durable Object, which manages real-time game state, player actions, AI behavior, and round progression. This allows multiple games to run concurrently while maintaining synchronized state across all players. The result is a system where voice AI, real-time infrastructure, and multiplayer interaction combine into something that feels less like a traditional game and more like a live social experiment.

Job searching is frustrating because most people are guessing. You submit the same application as everyone else and walk into interviews without really knowing what the company is looking for. Prelume changes that by generating a Signal Packet before you apply. It analyzes the role, the company, and the broader market to surface what actually matters, including hidden expectations, positioning gaps, and how to stand out as a candidate. Instead of generic advice, you get a clear strategy tailored to the role. On top of that, Prelume includes a live AI Career Advisor powered by ElevenLabs ElevenAgents. You can ask questions about the role, your positioning, or how to approach interviews, and get responses in real time through both text and voice. The agent uses the Signal Packet as context, so it feels like a focused, personalized coaching session rather than a generic chatbot. To generate these insights, Prelume uses Firecrawl to extract structured data from job postings and company signals. This makes it possible to move beyond surface-level descriptions and turn raw information into something actionable. The result is a tool that helps people understand what they are walking into, position themselves more effectively, and approach interviews with clarity instead of guesswork.