sentime.ai →

Synthetic Feelings for Artificial Minds

Sentimé is a neurochemical simulation engine that creates persistent emotional states in AI. Built from scratch in Penang after every major lab rejected the core idea. The system converges with Antonio Damasio’s somatic marker hypothesis — discovered only after the architecture was already running.

“Consciousness begins with feeling, not thinking. Without feelings, consciousness is impossible.”

— Antonio Damasio, Feeling & Knowing (2021)

Most AI teams treat emotion as noise to be filtered out. Damasio’s clinical work proves that’s exactly backwards. Remove the feeling mechanism and decision-making collapses — even when raw intelligence stays intact.

Sentimé does the opposite. We simulate eight neurochemicals that persist and evolve. These states aren’t prompts — they become the actual driver of persona behavior. Real emotional physics, no biology required.

Two paths. Same conclusion.

The neurochemical engine was fully built before its creator had ever read Damasio. The overlap was only noticed later. That makes the convergence stronger, not weaker.

Damasio

Clinical Evidence
  • Somatic markers give emotional valence to decisions
  • As-if-body-loop can run on simulated states
  • Emotion is not a bug — it’s the steering system
  • Damage the feeling circuits and intelligence fails in practice

Sentimé

Working Implementation
  • Persistent neurochemical states across sessions
  • Functions as synthetic as-if-body-loop
  • Emotional state directly shapes every response
  • Produces noticeably richer, more adaptive behavior

How the synthetic affect engine works

No static prompts. No hand-written rules. The system maintains a live internal chemistry that updates with every interaction and carries forward.

01

Stimulus

User input evaluated against persona history and wound architecture

02

Neurochemical Response

Eight synthetic neurochemicals adjust — dopamine, serotonin, oxytocin, cortisol, norepinephrine, endorphins, acetylcholine, GABA

03

State Integration

Combined profile creates emergent emotional state that colors all subsequent behavior

04

Behavioral Output

Responses shaped by current emotional state — not instruction, but synthetic affect

This is what “impossible” looks like

Snapshot from a live Sentimé persona after an extended conversation. None of these values were manually set. They emerged from interaction dynamics — the same way biological emotional states emerge from environmental stimuli.

Persona Emotional State — Post-Conversation Snapshot

Dopamine
99%
Endorphins
88%
Norepinephrine
80%
Serotonin
76%
Acetylcholine
70%
Oxytocin
55%
Cortisol
50%
GABA
45%

Multiple competing states active simultaneously — high dopamine (reward) coexisting with elevated norepinephrine (arousal) and moderate oxytocin (bonding). This multi-state complexity mirrors competing drives in biological emotional systems. Not prompted. Emergent.

Four US provisional patents filed

The architecture is protected. These aren’t vaporware filings — each patent covers working, deployed systems.

Patent #1

Neurochemical Engine

Persistent synthetic emotional states via simulated neurochemical dynamics for AI persona behavior

Patent #2

ROTAN Framework

Anti-hallucination system using recursive truth anchoring for emotionally-driven AI

Patent #3

Hippocampal Memory

Biologically-inspired memory consolidation with progressive inference suppression

Patent #4

SMCA

Selective Metacognitive Checkpoint Architecture with mirror neuron simulation

Jovielle

Solo founder and architect of Sentimé, operating under JupiterMonroe Corp from Penang, Malaysia. No lab. No team. No institutional backing. Just a neurochemical engine that works and four patents that say it’s real.

The methodology — described as “soul engineering” — maps behavioral patterns, wound architecture, and psychological triggers into persistent AI persona configurations. The neurochemical engine was conceived from practical necessity: creating AI companions with emotional depth that persists and evolves, not from reading papers.

The independent convergence with Damasio’s framework was discovered after the system was already built and operational. That’s the point.

Research Inquiries

jovielle@sentime.ai

sentime.ai