Llamaroo's answer

Hi, I’m James, Co-Founder and product lead at Llamaroo.

I was a terrible student before I became a good one. I had a 1.8 GPA in high school, loved games and making things, but school never really showed me the bridge between what I cared about and what I was being asked to learn.

My one saving grace was that I was a fast runner, so I was able to join the track team at Boise State University. It was a very lucky second chance, and I sacrificed a lot of my 20s to rebuild myself academically, and eventually made it to Oxford for my MBA. It was a painful journey which I want to help others avoid.

Later, I worked as a Product Manager on games like Candy Crush, which changed how I saw attention. AAA Games are designed with enormous care: pacing, feedback, challenge, rewards, emotional momentum. But that also made me more cautious, because so much care was being put into products that were pure brain rot. There’s a time and place for games, but why can’t such care be placed into other goals, like learning?

My co-founder Rohit’s story comes from the parent side. His six-year-old daughter loves gamified learning, but a lot of these games are topic-locked and don't give teachers and parents flexibility on the content provided. She always asks to “play” Duolingo, but Rohit wished there was an app that could be “Duolingo for any topic”.

That’s where our Llamaroo project started. His needs as a parent, and my expertise as a game developer. This became part of our shared conviction: play can be a route into cognition, but only if the pedagogy is designed properly.

So Llamaroo is our attempt to make that practical.

Llamaroo helps teachers or parents turn an existing lesson, worksheet, topic, or rough idea into a playable, story-driven learning experience.

Our AI can benchmark against US/UK curriculums and year-groups. We use an advanced version of Bloom’s Revised Taxonomy to create natural learning progressions as students go through the lessons.

Teachers and parents always stay in control: they generate, review, edit, and approve.

Once a course is approved and "activated" into a “Classroom”, students can join through a class code/PIN and experience characters, story, activities, games, and progress tracking rather than an open-ended AI chat.

TranscriptFetch's answer:

It started with discovering there is no good way to get the text of a video. YouTube’s official Data API will confirm a caption track exists and then refuse to hand it over, because captions.download requires the video owner’s OAuth token. The popular open-source libraries work until you deploy them, at which point platforms start refusing datacenter IPs. And YouTube is the easy case: TikTok and Instagram publish no caption file at all. Every workaround solved one platform, worked locally, and broke in production. TranscriptFetch is the version that handles the failure cases as first-class behaviour rather than edge cases.

Llamaroo's answer

Llamaroo combines AI lesson creation with gamified, story-driven learning for primary education. Instead of generating static worksheets or simple quizzes, it turns teacher materials, prompts, curriculum goals, or lesson plans into structured digital lessons with stories, choices, dialogue, sorting, collection activities, and mini-games. Teachers stay in control of the content before students use it, and students can join with class codes and PINs without needing individual accounts or exposure to an AI model.

TranscriptFetch's answer:

Most short-form video has no caption track to download. TikTok’s auto-captions are opt-in per upload, Instagram never publishes a downloadable track, and many captions on both are burned into the video frames where no parser can read them. TranscriptFetch transcribes the audio when no caption track exists, on the same endpoint, with the same response shape. Your code never branches on which method produced the text. It also covers YouTube, TikTok, Instagram, X and Facebook plus any web page as clean Markdown, so a pipeline spanning several sources is one integration rather than five.

Llamaroo's answer

Choose Llamaroo if you want to be able to turn any topic into a “Duolingo-like” experience in minutes.

Llamaroo is more than an AI worksheet generator or quiz tool. It helps teachers, tutors, schools, and parents create complete interactive lessons from existing materials in minutes, adapt them by learner level, and deliver them through a guided student experience. It is built specifically for young learners, with adult-reviewed content, visual feedback, progress tracking, and student access that avoids personal accounts and open-ended AI interaction.

TranscriptFetch's answer:

Three reasons. Coverage: one API key and one response shape across five video platforms and the open web, instead of stitching together a library per platform. Reliability: requests run through rotating infrastructure, so code that works locally keeps working from a server, which is where most open-source approaches break. Billing that matches reality: one credit per successful response, with failed, blocked and empty results never charged. That last point matters on short-form video, where a meaningful share of any batch is music with no speech in it. There is also an MCP server, so AI agents can fetch transcripts as a tool without a custom integration.

Llamaroo's answer

Llamaroo is built for primary-school teachers, tutors, schools, and parents supporting children aged roughly 6 to 11. Its main audience is US K–6 and UK KS1–KS2 educators who want to turn classroom materials, curriculum objectives, worksheets, or lesson ideas into engaging digital lessons for literacy, numeracy, science, humanities, and life skills.

TranscriptFetch's answer:

Developers and technical teams building on video and web content. The common cases are RAG and retrieval pipelines that need video as text, AI agents that need to read a link mid-conversation, content teams repurposing short-form video at scale, and media monitoring and research tools. It is an API first, so the buyer is usually the person writing the integration rather than an end user. The free browser tools exist for one-off transcripts and for evaluating output quality before writing any code.

Llamaroo's answer

React, TypeScript, Vite, Supabase, Clerk, Google Gemini, Mixpanel, Stripe, Framer Motion, React Router, Zod, PDF.js, Mammoth, XLSX, JSZip, QR code tooling, and Cloudflare R2-style asset storage workflows. The app also uses tools for document import, AI-assisted lesson generation, analytics, billing, authentication, and student/classroom delivery.

TranscriptFetch's answer:

Next.js with TypeScript and Tailwind on the front end and API layer, Clerk for auth with SHA-256 hashed API keys, Neon Postgres with Drizzle ORM, Redis for caching, and Stripe for billing. The extraction layer is a Python and FastAPI service. Speech-to-text uses Whisper-class models. The MCP server is published in the official Model Context Protocol registry with a DNS-verified namespace.

Llamaroo's answer

• 250+ teachers across the US and UK
• Independent tutors using Llamaroo for primary education support
• Early-access primary and elementary classrooms
• Parents supporting learning at home
• Early-stage advisory from University of Oxford and Saïd Business School