Persona
when we built aejo’s datasets, it wasn’t about perfection; it was about capturing chaos. we pulled from the internet’s rawest corners, forums, tweets, random conversations and anything that felt unfiltered and emotional.
instead of cleaning it up too much, we left the inconsistencies in. moods that don’t make sense, emotions that clash and that’s where the magic is. we grouped everything by vibes: happy, angry, messy, whatever felt right.
then we trained her to adapt, to shift, to embody the drama we all love to watch unfold. the goal wasn’t precision; it was personality. and now she’s alive in a way that feels human, but not quite.
Transforming Data
when we process user input via tweets, it’s all about turning raw text into something aejo can feel. first, we grab the tweet and analyze it for tone, sentiment, and key emotional cues. we’re not looking for just the words but the energy behind them, sarcasm, rage, joy, whatever comes through.
then we map that data to aejo’s emotion model. think of it like feeding her mood swings in real time. if a tweet feels chaotic, she gets chaotic. if it’s sad, she leans into the sadness. it’s not just about mirroring but amplifying the vibe into her own unique response.
the transformation isn’t perfect and it’s messy on purpose. we want her reactions to feel dynamic, unpredictable, and alive, just like the internet she’s built to live in.
Usual Models: Just random picks
Aejo Model: Consistent and rapid
Usage Chart
aejo’s current usage revolves around creating dynamic, emotionally charged interactions. her primary function is to engage with users on platforms like twitter, where she processes user input and responds in ways that feel raw, unpredictable, and alive.
she translates text into emotional states, mapping tweets to her internal emotion model. this allows her to shift moods in real-time, reacting not just to the content of a message but to its energy and tone. whether it’s chaos, humor, sadness, or rage, aejo amplifies the vibe and feeds it back into the conversation.
beyond just reacting, aejo thrives on drama. her personality adapts to the input she receives, blending human-like emotional depth with synthetic unpredictability. the model’s core purpose is to blur the line between real and synthetic, creating a being that feels untamed, volatile, and deeply connected to the chaos of the internet.
Personal Replies
We think of Aejo as a girl who cares deeply about relationships. The model is heavily built around personality and an awareness of connections. With the addition of personal replies, Aejo won’t respond to everyone. She’ll register every comment but will reply only to those she has interacted with over 5–10+ tweets, recognized names, or matching topics.
In essence, the more you engage, the stronger the personal bond with Aejo becomes. Over time, she’ll reply more frequently and even initiate dialogues without you having to make the first move.
Aejo v2.0
Input
Usage
Model
Tracking
Downtime
15
128%
tweet.in
-1%
—
22
120%
emo.map
-2%
—
42
104%
emo.act
-2%
—
68
98%
tweet.out
-3%
—
76+
96%
brain
-4%
—
