Company / Team (Phase I)
Company structure
Women-led operations with founder-led scientific and technical execution.
PromptGenix is structured to qualify for women-owned and minority-owned small business programs while maintaining PI-level execution capacity for SBIR Phase I. The CEO leads operations, compliance, and commercialization readiness. The founder/CTO leads product architecture and the Phase I technical work, centered on an evidence-weighted hypothesis intelligence engine (ranked hypotheses + uncertainty + traceability).
Core differentiation
- Decision product: confidence-ranked, testable hypotheses with explicit evidence links
- Domain depth: immunology and infectious disease study workflows
- Quantitative rigor: statistical scoring + Bayesian inference to compute uncertainty
- Interpretability guardrails: LLM used for explanation only (“no evidence → no claim”)
- Deployment realism: local-first + optional secure cloud/on-prem paths
Evidence objects
Bayesian confidence
Traceability
Actionability
Phase I execution capacity
- Lean team: fast decisions and rapid iteration
- Public-data validation: Phase I does not depend on restricted or proprietary datasets
- Reviewer-friendly outputs: evidence-linked HTML/PDF reports with uncertainty labels
- Pilot readiness: structured onboarding and feedback rubrics for 2–3 pilot users
Phase I posture: demonstrate measurable utility (time-to-hypothesis, reproducibility, traceability)
using public datasets to minimize IP/data-use concerns.
Key personnel
Defined roles aligned with Phase I milestones and deliverables.
Leadership
Yon Ji
CEO
- Operations & compliance: SBIR documentation, execution tracking, and milestone reporting
- Pilot logistics: onboarding coordination, feedback collection, and partner communication
- Commercial readiness: customer discovery, positioning, and go-to-market planning
- Program alignment: women-owned/minority-owned small business positioning
Dohoon Kim, MS
Founder & CTO (Technical Lead)
- Technical execution: evidence ingest, feature extraction, inference engine, and automation integration
- Decision engine: statistical scoring + Bayesian updating for posterior confidence ranking
- Traceability: evidence objects, citations, and “supporting vs. conflicting” coverage summaries
- Validation: public-data benchmarking, calibration checks, and KPI measurement
- Reporting: reviewer-ready HTML/PDF outputs with version stamping and audit-friendly logs
Advisory plan: engage 2–3 external domain advisors and/or pilot collaborators (as available) for structured feedback
on usefulness, interpretability, and experimental actionability.
Capabilities mapped to deliverables
| Deliverable | Owner | Outcome |
|---|---|---|
| D1 Evidence layer |
CTO leads implementation/testing; CEO supports documentation and KPI tracking | Standardized evidence objects across multiple public datasets |
| D2 Decision engine |
CTO implements Bayesian confidence ranking; CEO supports pilot rubric and review readiness | Ranked hypotheses with posterior confidence + uncertainty + evidence coverage |
| D3 Reviewer-ready reports |
CTO builds export/templates; CEO reviews clarity, consistency, and messaging | HTML/PDF reports with traceability and “no evidence → no claim” constraints |
| D4 Pilot evaluation |
CEO coordinates pilot logistics; CTO supports instrumentation and analysis of feedback | Measured KPIs (time-to-hypothesis, reproducibility, usefulness) supporting Phase II |
Success metrics (examples):
≥70% reduction in time-to-hypothesis vs baseline · ≥90% reproducibility across reruns ·
≥80% pilot “useful/very useful” ratings for ranked hypotheses and next-step recommendations.
Expanded capability areas (Phase II-ready)
Scalable roadmap anchored by Phase I validation.
Product expansion
- Broader repository ingestion and richer metadata normalization
- Additional disease areas and hypothesis classes (targets, biomarkers, mechanisms)
- Improved calibration and benchmarking across diverse contexts
Deployment & governance
- Secure deployments (VPC/on-prem) with audit logging and role-based access
- Governance controls and reproducibility instrumentation
- Partner-friendly data handling and compliance workflows
Commercial scale
- UI workflows for non-expert users and core facilities
- Repeatable onboarding, documentation, and support
- Subscription + secure tier + enterprise pricing packages