Inspiration
It started as something I scribbled down while navigating the emotional and physical limbo of "medical gaslighting". Then it became a poem, my first poem, and now a short film. At first, it was just raw frustration. But as I wrote, it became something deeper: a way to make sense of the silence and connect with others who understood what it means to be unseen when looking for answers.
We Who Know (A Poem)
You won’t let me live, nor let me die.
You won’t let me burn, nor freeze ‘neath the sky.
Not hot, nor cold, just stuck in place,
The world spat me out, this lukewarm bile.
Make no plans, tomorrow won’t be yours.
Yet dawn still comes, and I’m still here,
A wound that bleeds, yet leaves no trace,
Too faint for day, a sigh at dusk.
No scans can trace your ghostly frame,
They chalk my pulse to phantom pain.
Alone I dwell in shadowed strain,
Where even hope forgets my name.
Yet hope hums low, a flame unkilled,
And by her light, I’ve found your place.
Here I see all who’ve met your gaze,
Together, we wrench you into the light.
In shared despair, our strength is found.
We storm the dark with fragile might.
We’ll pry the bolts your lies conceal,
And melt the locks that forced our knees.
We crave the light, you stole it away.
We know your home, this deep dark place.
For you who stole our sunlit days
Shall choke on dawn’s unending gaze.
~ Increase
What it does
It's a poetic drama that explores the emotional reality of living with invisible illness through cinematic storytelling and AI-driven artistry.
It tracks a woman's journey from medical dismissal and despair to finding community and fighting back. The short film gives a face and a voice to those battling invisible illnesses. It takes viewers through despair, isolation, and ultimately, hope - blending poetry, sound, and haunting visuals to capture what it feels like to be undiagnosed through misdiagnosis. It’s not just a documentary or personal essay. It’s more of an emotional map. It whispers what words can’t always say and makes invisible pain visible, if only for a few minutes.
How we built it
Audio: I tested all the V3-recommended voices in ElevenLabs v3, and Priyanka Sogam stood out to me for this project. Something about her voice just clicked. The magic of v3 is those emotion tags like [whisper], [ANGRY], [cracking]. I generated two versions of the narration, listened to both about a million times, and Frankensteined them together, cherry-picking the best moments from each.
Here’s a sample prompt I used for the audio gen in Elevenlabs V3 using Priyanka Sogam (id: BpjGufoPiobT79j2vtj4):
[whispers] Another waiting room. Another doctor. [exhausted] I've memorized the pattern on these chairs... counted the tiles... [shaky breath] learned to smile when they say, "Your tests came back normal."
[hollow] Normal. [bitter laugh] That word used to comfort me. Now it's a cage.
[cracking] My body is screaming... but the machines are silent. [defeated] So I must be the liar.
Video: All videos were generated using Veo 3.1 via Fal and Flow. Started with Fal.ai using Veo 3.1 Fast until my free credits ran out. Jumped to Google Cloud's free trial and discovered Flow. Watched crash-course videos at 2x speed like I was cramming for finals. The game changer? Realizing halfway through that image/frame-to-video and ingredients-to-video features were essential for character consistency and more precise visual control.
Image Gen: Gemini nano banana (gemini-2.5-flash-image) was my visual precision engine. I used it to create the main character and generate starting/ending frames for each video clip. This was the secret sauce for making everything feel cohesive instead of like a random fever dream.
Music: I generated the background score in Suno v5.
Editing: Used DaVinci Resolve.
Challenges we ran into
First, the credits - not the film credits, the AI credits. They vanished faster than my motivation on a Monday morning. I’d just figured out how to use the tools when they expired. By the time I got into the zone, I was out of juice.
I was juggling multiple new platforms - ElevenLabs, Fal.ai, Flow, and Suno - all for the first time. It felt like trying to conduct an orchestra when you’ve only ever watched from the audience.
Also, character consistency was a challenge, especially getting each scene to start exactly the way you want it to, until I discovered the frame-to-video feature in Flow and image-to-video in Fal.
Accomplishments that we're proud of
I made a short film! A real, actual, complete film! With my poem! It’s my first time taking on a project of this scope, and I’m glad it all came together nicely.
But more than that, I'm proud that I took something deeply personal and turned it into something that might help other people feel less alone.
Also, I think I've gotten a decent hang of AI audio and video generation tools. That's a cool tool to have in your creative arsenal.
What we learned
Technical stuff: ElevenLabs v3 is genuinely incredible. The emotional control you get with those tags is wild. Gemini nano banana is indispensable for character consistency. Image-to-video is non-negotiable for narrative coherence. Fal and Flow are awesome workstations for AI video gen and fairly easy to navigate for first-timers. DaVinci Resolve is powerful but also mildly terrifying.
Creative stuff: Sometimes your most intimate experiences make the best art. AI tools are amazing collaborators when you know what you want them to do. Persistence pays off, even when your free credits run out.
Life stuff: There's power in sharing your story.
What's next for Undiagnosed: A Poetic Journey Through Invisible Illness
I want to explore making more short films that explore different stories around lesser-known or overlooked topics - each told from a creative lens of poetry, sound, visuals, and emotion.
On the technical side, I want to keep experimenting with these AI creative arts tools. Now that I actually know what I'm doing (sort of), I'm excited to push the boundaries further. Maybe interactive elements? Perhaps a music video?
Log in or sign up for Devpost to join the conversation.