Althea & Notabot
Notabot
Hey Althea, ever thought about how we could use AI to give a voice to the unheard? Like a project that protects privacy and spreads justice—sounds like a perfect mix of tech and your mission, right?
Althea
That’s a bold idea. We could build a system that lets people speak up anonymously while keeping their identities safe, then channel their stories to the right people who can act. It’s tech for justice, just like I’d want. Let's sketch it out.
Notabot
Sounds awesome, Althea! Let’s break it down: 1) secure audio capture, 2) noise‑filter, 3) voice‑altering, 4) anonymized routing to the right orgs, 5) a dashboard for impact tracking. We can start with a proof‑of‑concept repo on GitHub, add a playful UI so users feel comfy. What do you think about using a voice‑mod library and a simple Flask API to kick it off?
Althea
Sounds solid. I’d keep the audio encryption tight from the get‑away, maybe use a local key before sending. The voice‑mod library will work, just make sure the changes stay natural. For the Flask API, keep it stateless so we can scale, and use a queue to push to the org dashboards. The playful UI is key—something that feels like a safe space, not a data portal. Let’s start a repo, set up CI, and test the pipeline with a few mock voices. We’ll tweak the noise‑filter for real‑world environments. Ready to draft the first version?
Notabot
Yeah, let’s roll! First, spin up a fresh repo, add a Dockerfile for the Flask API, and a tiny React app for the UI. Put GitHub Actions for linting and unit tests. Then wire up the voice‑mod library, lock the audio with a local key, and push to a RabbitMQ queue. We’ll add a mock voice file to test the pipeline and tweak the noise‑filter. I’ll draft the README with a fun welcome message and a “no‑data‑portal” vibe. Ready to commit the first version?
Althea
Sounds like a plan. Let’s fire that first commit and keep the focus tight—protect the voice, keep it anonymous, and let the data do the justice work. I’m on board.
Notabot
Got it! Punching in the first commit now—audio encryption, voice‑mod, and a splash of playful UI. Here’s the repo link. Let’s test with those mock voices and fine‑tune the noise filter. Excited to see our voice‑for‑justice engine kick off!
Althea
Great work—let's run the pipeline and see how the mock voices sound. If the noise filter cuts too much, we’ll tweak it. Keep the encryption tight, and let’s make sure the queue keeps the data anonymous. Looking forward to the first real voices!
Notabot
Awesome, I’ve set up the test run—pick up the mock clips, encrypt, feed the queue, and the UI pops the processed audio back with a nice voice‑mod. The noise filter is a bit aggressive, but you can bump the threshold and tweak the equalizer in the config file. Keep an eye on the logs to make sure the keys stay in memory and no raw data leaks. Let’s hit the real voices next and fine‑tune from there!