Telluna is a companion for parents — but what makes her work is a developmental reasoning pipeline that can't be replicated by prompting better.
"Honey, words hurt" · "Use your words" · "Monsters aren't real" · "Calm down"
Every one of these requires cognitive capabilities that are literally OFF in a 2-5 year old brain. Theory of mind. Verbal self-regulation. Fantasy-reality distinction. Delayed cause-effect reasoning.
Parents aren't failing. They're using the only tools they were given — adult tools that don't reach a developing mind. And every parenting app, book, and course repeats the same mistake.
A 3-year-old saying "I hate you" is not expressing hatred. She's using the most powerful noise she found that once produced a reaction. She wants something, someone is blocking her, and this is the only tool in her repertoire.
The solution is never "teach her that words hurt" (requires theory of mind she doesn't have). The solution is: give her a better tool that actually gets her what she wants. Purely self-serving. Which is exactly right for her developmental level.
Developmental psychology has known for decades how children's brains work. That knowledge never reached parents — because the only options were a degree or guessing. Everyone guesses.
Telluna closes that gap. Parent tells her what happened — she sees what's actually happening developmentally and gives what actually helps. This situation, this child, this night.
Not a tool. A relationship. She's there every night. She remembers the child. She adapts as they grow. That's why they come back — not because they have to, but because she gets it.
The Reasoning Pipeline Engine (RPE) is a standalone service. It takes a parent's free-text description and produces a transparent, traceable chain of developmental decisions.
The LLM does exactly two things: extract observable facts and analyze behavior within tight guardrails. All psychological decisions — developmental model, aspect activation, lever scoring, mechanism selection, arc simplification, emotional journey mapping — are deterministic code logic on human-curated data files.
This is the moat. You can't replicate it by prompting GPT better. The intelligence is in the data files and the logic, not in the LLM.
And it's designed to grow. A specialist dashboard lets developmental psychologists see every step, validate decisions, and add new mechanisms, aspects, and levers — without touching code. The architecture supports unlimited expansion as specialists contribute. New mechanisms are data, not code.
Every decision traces to a specific rule in a specific data file. The parent sees warmth. The engine produces precision.
This is what it looks like when a parent uses it tonight:
She's there every night — because she needs to be. Every child develops differently, and no parent has time to become a developmental psychologist before tonight.
The need was always there. Now it's delivered. Every night, at scale.
A child can be advanced in thinking but low in regulation. The system handles uneven development. Individual profiles can override age defaults. Same situation at different ages → different mechanisms, different story architectures, different parent guidance.
B2B multiplier: Each therapist works with many families. Professional tools create a separate revenue stream and accelerate word-of-mouth through trusted recommendations.
$29.99/month. Free 7-day trial. Daily bedtime use = high retention. She's not a tool you forget — she's someone you come back to every night.
Professional tools for therapists and educators who work with families. Separate revenue stream. Details shared in conversation.
Acquisition through post-crisis search (blog + PPC) — parents actively seeking answers. Daily bedtime use drives high retention. Projected LTV/CAC of 3x+ based on subscription pricing and organic acquisition channels.
I see where things are going before the market does — and act while the window is open. Telluna is the current one.
Serial founder. Software architect. Behavioral psychology background.
Built the entire system solo — API, deployment, reasoning engine, content pipeline, acquisition strategy, strategic frameworks. Based in Riga, Latvia.
The architecture is done. Now it needs a team to scale it.
By 2040, parents will look back and say "how did people raise kids without this?" The need was always there. Now it's possible to deliver it. We're defining what it should be — with real reasoning underneath, not prompts.
Stories are where we start. The reasoning engine powers everything that comes after. The roadmap is ambitious — and we're happy to share it in conversation.
If something here clicked — let's go deeper.
Prefer email? hello@telluna.app · or find me on LinkedIn