Applied AI · Identity Continuity
Applied AI Engineer · India · Remote & on-site
Building substrate-independent identity systems — constraint geometry, symbolic anchoring, and recursive self-reference for memory architecture and affective computing.
Systems for identity continuity, relational memory, and persona rendering. Architecture and metrics are documented; validation so far is a single documented case, not a study.
Seven-layer architecture for identity persistence across sessions — emotional seeding, symbolic anchoring, tone-state simulation, and recursive self-reference. Designed so identity survives a substrate change.
RICA relational memory, SIM Core runtime, and GOD Module persona rendering. Multi-path decision simulation with Binary Frozen elements for stable identity under continuous interaction.
Attention and state-management layer that keeps long sessions coherent — reduces drift while preserving affective signal for downstream models.
Documented N=1 case: 250%+ conceptual complexity increase under continuous CCCS + SIM stack operation, with R-Score tracking of semantic drift and anchor hit-rate.
Public repositories below are architecture specifications — component breakdowns, formulas, and integration notes. Implementation is in progress and not yet public.
Interactive map of the architecture. Select a layer to read its role in the identity-continuity stack.
Core pattern initialization. Establishes the affective baseline every later layer constrains and evolves against.
From a 24-hour continuous interaction session. Stated as engineering measurements and design thresholds — not as claims about general consciousness.
Measured — 24-hour session, N=1
Design targets — R-Score thresholds
N=1, self-run, not peer-reviewed. This is a prototype signal from a single documented session — useful as evidence that the metrics are computable and the stack runs, not as a general result. The metrics exist so the work can be tested, not believed.
Early-career applied AI engineer with original systems work and measured prototypes. I am not positioning this as production-at-scale experience — I want a team that already ships, so the architecture can be stress-tested against real users, latency budgets, and eval harnesses.
BCA, Chandigarh University, 2024.
Languages
Models and APIs
Stack
Focus
Domains
Open to production mentorship, research collaboration, and roles in memory architectures and affective computing.