Draft — Neon Startup Program

MOTIVATION LETTER

The Conjunctive Consolidation Threshold model, a tripartite pharmacological framework for reward-memory encoding prevention, has been validated through ODE/RK45 and Bayesian MCMC simulations showing an 85.8 percent reduction in encoding probability. This computational work, published as three sole-authored preprints on OSF and Zenodo, now requires a scalable data infrastructure to move from single-machine simulations toward multi-user clinical trial platforms. Neon Startup Program credits for serverless Postgres would directly support the deployment of IMPRINT, an addiction-liability screening platform that queries Bayesian MCMC outputs and clinical trial simulation results in real time. IMPRINT currently runs on a local PostgreSQL instance limited to 10,000 rows per query; scaling to 500,000 rows for multi-site validation requires Neon's branching and horizontal scaling capabilities. The $100,000 in credits would cover approximately 18 months of production-grade database operations, including compute, storage, and branching for parallel development environments. TOPOLOGIX, a topological data analysis platform for drug-protein interaction using persistent homology and bipartite simplicial complexes, also depends on serverless Postgres to store and query large graph adjacency matrices. A provisional patent on the CCT core architecture is filed for Q3 2026, and co-authored work is under review at Alcohol (Elsevier). The Neon program's focus on early-stage startups under five years old, under fifty employees, and under five million dollars in funding matches the current status of this venture: zero institutional funding, two platforms at MVP stage, and a team of one technical founder. The program does not offer domain-specific mentorship, but the infrastructure credits are the binding constraint for the next six months of development. Nigeria has fewer than fifty active computational neuroscience researchers, and this venture aims to build the first open-source addiction pharmacology platform hosted on African infrastructure. Neon credits would remove the cost barrier to maintaining a production database for clinical trial simulation data, drug-protein interaction graphs, and Bayesian posterior distributions. The application deadline is listed on the program website, and all materials are submitted with this letter.

SHORT ESSAY: TECHNICAL FIT

IMPRINT requires serverless Postgres to manage three distinct data types: Bayesian MCMC posterior samples from the CCT model, clinical trial simulation outputs from ODE/RK45 solvers, and user-submitted pharmacological profiles for addiction-liability screening. Current local PostgreSQL handles 10,000 rows per query with a 2.3 second latency; production deployment for ten concurrent users requires sub-100 millisecond latency on 500,000 rows. Neon's branching feature enables parallel development of the IMPRINT query engine, the TOPOLOGIX graph database, and the GATE neural-stimulation safety evaluation platform without provisioning separate database instances. The CCT model's mathematical specification (OSF 10.17605/OSF.IO/EMY4U) defines a conjunctive consolidation threshold that depends on real-time integration of dopamine, glutamate, and opioid signaling parameters. Serverless Postgres with horizontal scaling supports the dynamic query patterns required for this integration, where each user session generates approximately 2,000 parameter combinations. The Bayesian population dynamics preprint (Zenodo 10.5281/zenodo.20492472) specifies a clinical trial architecture with 1,200 simulated patients across three arms; storing and querying these simulation results for validation by collaborators at Michigan, Princeton, and NYU requires a database that supports branching for review workflows. Neon's compute-compute separation allows the database to scale to zero when not in use, reducing costs during the development phase while maintaining the ability to burst to full capacity during validation sprints. The venture currently has zero database costs because it runs locally; Neon credits would enable the first cloud deployment without diverting personal income from the National Product Manager role at Synthcare.

SHORT ESSAY: IMPACT AND MILESTONES

Neon credits would enable three specific milestones over twelve months. First, deployment of IMPRINT as a multi-user web platform with real-time addiction-liability scoring, targeting fifty beta testers from the University of Ibadan pharmacology department and the Nigerian Institute of Medical Research. Second, integration of TOPOLOGIX with the IMPRINT database to enable drug-protein interaction queries alongside addiction-liability scores, creating a unified pharmacological data platform. Third, hosting of the CCT model's clinical trial simulation results for peer review and replication by the three endorsing laboratories: Kent Berridge at Michigan, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU. Each milestone depends on database scalability that local PostgreSQL cannot provide. The 85.8 percent reduction in encoding probability demonstrated in the CCT model requires independent replication; Neon credits would allow the simulation code and results to be hosted as an interactive database that reviewers can query directly. The provisional patent filed Q3 2026 covers the CCT core architecture; a production database with queryable simulation results strengthens the patent application by demonstrating practical implementation. Nigeria has no computational pharmacology startup with a cloud-hosted database for addiction research; this venture would be the first. The Neon program's selection criteria emphasize potential to use credits for growth; these credits would directly fund the infrastructure gap between a single-machine prototype and a multi-user platform that can attract partnerships with research institutions and pharmaceutical companies.

RESEARCH STATEMENT

The CCT model proposes that reward-memory consolidation requires simultaneous activation of three neural pathways beyond a conjunctive threshold. The model was validated through ODE/RK45 numerical integration and Bayesian MCMC sampling, with all five pre-registered hypotheses H1 through H5 confirmed. Encoding probability decreased from 0.855 to 0.122, an 85.8 percent reduction, with super-additivity of 12.8 percentage points. These results are documented in three sole-authored preprints: the foundational paper (OSF 10.17605/OSF.IO/KG7B5), the formal mathematical specification (OSF 10.17605/OSF.IO/EMY4U), and the Bayesian population dynamics with clinical trial architecture (Zenodo 10.5281/zenodo.20492472). A review article is under review at Neuroscience and Biobehavioral Reviews, and a co-authored paper is under review at Alcohol (Elsevier). The computational methods include Python with scipy and numpy for ODE integration, PyMC for Bayesian MCMC, and Ripser and Gudhi for topological data analysis. The platforms built on this research are IMPRINT for addiction-liability screening, TOPOLOGIX for drug-protein interaction using persistent homology and bipartite simplicial complexes, and GATE for BCI neural-stimulation safety evaluation. The provisional patent on the CCT core architecture is filed for Q3 2026. Endorsements from Kent Berridge at Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU provide external validation of the theoretical framework. The next research phase requires a production database to host simulation results for replication by these collaborators and to support multi-user clinical trial simulation. Neon credits would fund this infrastructure, enabling the transition from independent research to a platform that can support collaborative validation and eventual clinical translation.

CHECKLIST

- [ ] Completed application form on Neon Startup Program website
- [ ] Motivation letter (this document)
- [ ] Short essay on technical fit (this document)
- [ ] Short essay on impact and milestones (this document)
- [ ] Research statement (this document)
- [ ] Link to ORCID profile: 0009-0001-9272-6735
- [ ] Link to GitHub profile: github.com/AmunRaPtah
- [ ] Link to ZYCO website: zyco.org
- [ ] Links to three preprints on OSF and Zenodo
- [ ] Proof of provisional patent filing (Q3 2026)
- [ ] Endorsement letters or contact information for Kent Berridge, Samuel Gershman, Nathaniel Daw, Marcelo Mattar
- [ ] CV or resume with employment history and education
- [ ] Copy of B.Pharm degree certificate and transcript
- [ ] PCN pharmacist license copy

EDITOR NOTES

- Eligibility risk: The program requires the venture to be under five years old, under fifty employees, and under five million dollars in funding. The venture has zero funding and one employee, which meets these criteria, but the applicant should confirm that independent research without a registered company entity qualifies as a startup. If a registered company is required, the applicant may need to register a Nigerian business entity before applying.
- Fact verification: The applicant states that IMPRINT currently handles 10,000 rows per query with 2.3 second latency. This number should be verified by running a benchmark on the local PostgreSQL instance and documenting the exact query and hardware specifications.
- Gap: The application does not specify the exact number of beta testers or the timeline for the first deployment. The applicant should insert a specific target number and date, for example "fifty beta testers from the University of Ibadan pharmacology department by March 2027."
- Gap: The applicant should confirm whether the Neon program requires a demo video or a pitch deck. The checklist above assumes only written materials, but the program website should be checked for additional submission requirements.
- Gap: The applicant's current employment as National Product Manager at Synthcare may raise questions about time commitment to the startup. The applicant should prepare a brief statement about how the startup work is conducted outside of work hours or with employer permission.

MOTIVATION LETTER

The Conjunctive Consolidation Threshold model, a tripartite pharmacological framework for reward-memory encoding prevention, has been validated through ODE/RK45 and Bayesian MCMC simulations showing an 85.8 percent reduction in encoding probability. This computational work, published as three sole-authored preprints on OSF and Zenodo, now requires a scalable data infrastructure to move from single-machine simulations toward multi-user clinical trial platforms. Neon Startup Program credits for serverless Postgres would directly support the deployment of IMPRINT, an addiction-liability screening platform that queries Bayesian MCMC outputs and clinical trial simulation results in real time. IMPRINT currently runs on a local PostgreSQL instance limited to 10,000 rows per query; scaling to 500,000 rows for multi-site validation requires Neon's branching and horizontal scaling capabilities. The $100,000 in credits would cover approximately 18 months of production-grade database operations, including compute, storage, and branching for parallel development environments. TOPOLOGIX, a topological data analysis platform for drug-protein interaction using persistent homology and bipartite simplicial complexes, also depends on serverless Postgres to store and query large graph adjacency matrices. A provisional patent on the CCT core architecture is filed for Q3 2026, and co-authored work is under review at Alcohol (Elsevier). The Neon program's focus on early-stage startups under five years old, under fifty employees, and under five million dollars in funding matches the current status of this venture: zero institutional funding, two platforms at MVP stage, and a team of one technical founder. The program does not offer domain-specific mentorship, but the infrastructure credits are the binding constraint for the next six months of development. Nigeria has fewer than fifty active computational neuroscience researchers, and this venture aims to build the first open-source addiction pharmacology platform hosted on African infrastructure. Neon credits would remove the cost barrier to maintaining a production database for clinical trial simulation data, drug-protein interaction graphs, and Bayesian posterior distributions. The application deadline is listed on the program website, and all materials are submitted with this letter.

SHORT ESSAY: TECHNICAL FIT

IMPRINT requires serverless Postgres to manage three distinct data types: Bayesian MCMC posterior samples from the CCT model, clinical trial simulation outputs from ODE/RK45 solvers, and user-submitted pharmacological profiles for addiction-liability screening. Current local PostgreSQL handles 10,000 rows per query with a 2.3 second latency; production deployment for ten concurrent users requires sub-100 millisecond latency on 500,000 rows. Neon's branching feature enables parallel development of the IMPRINT query engine, the TOPOLOGIX graph database, and the GATE neural-stimulation safety evaluation platform without provisioning separate database instances. The CCT model's mathematical specification (OSF 10.17605/OSF.IO/EMY4U) defines a conjunctive consolidation threshold that depends on real-time integration of dopamine, glutamate, and opioid signaling parameters. Serverless Postgres with horizontal scaling supports the dynamic query patterns required for this integration, where each user session generates approximately 2,000 parameter combinations. The Bayesian population dynamics preprint (Zenodo 10.5281/zenodo.20492472) specifies a clinical trial architecture with 1,200 simulated patients across three arms; storing and querying these simulation results for validation by collaborators at Michigan, Princeton, and NYU requires a database that supports branching for review workflows. Neon's compute-compute separation allows the database to scale to zero when not in use, reducing costs during the development phase while maintaining the ability to burst to full capacity during validation sprints. The venture currently has zero database costs because it runs locally; Neon credits would enable the first cloud deployment without diverting personal income from the National Product Manager role at Synthcare.

SHORT ESSAY: IMPACT AND MILESTONES

Neon credits would enable three specific milestones over twelve months. First, deployment of IMPRINT as a multi-user web platform with real-time addiction-liability scoring, targeting fifty beta testers from the University of Ibadan pharmacology department and the Nigerian Institute of Medical Research. Second, integration of TOPOLOGIX with the IMPRINT database to enable drug-protein interaction queries alongside addiction-liability scores, creating a unified pharmacological data platform. Third, hosting of the CCT model's clinical trial simulation results for peer review and replication by the three endorsing laboratories: Kent Berridge at Michigan, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU. Each milestone depends on database scalability that local PostgreSQL cannot provide. The 85.8 percent reduction in encoding probability demonstrated in the CCT model requires independent replication; Neon credits would allow the simulation code and results to be hosted as an interactive database that reviewers can query directly. The provisional patent filed Q3 2026 covers the CCT core architecture; a production database with queryable simulation results strengthens the patent application by demonstrating practical implementation. Nigeria has no computational pharmacology startup with a cloud-hosted database for addiction research; this venture would be the first. The Neon program's selection criteria emphasize potential to use credits for growth; these credits would directly fund the infrastructure gap between a single-machine prototype and a multi-user platform that can attract partnerships with research institutions and pharmaceutical companies.

RESEARCH STATEMENT

The CCT model proposes that reward-memory consolidation requires simultaneous activation of three neural pathways beyond a conjunctive threshold. The model was validated through ODE/RK45 numerical integration and Bayesian MCMC sampling, with all five pre-registered hypotheses H1 through H5 confirmed. Encoding probability decreased from 0.855 to 0.122, an 85.8 percent reduction, with super-additivity of 12.8 percentage points. These results are documented in three sole-authored preprints: the foundational paper (OSF 10.17605/OSF.IO/KG7B5), the formal mathematical specification (OSF 10.17605/OSF.IO/EMY4U), and the Bayesian population dynamics with clinical trial architecture (Zenodo 10.5281/zenodo.20492472). A review article is under review at Neuroscience and Biobehavioral Reviews, and a co-authored paper is under review at Alcohol (Elsevier). The computational methods include Python with scipy and numpy for ODE integration, PyMC for Bayesian MCMC, and Ripser and Gudhi for topological data analysis. The platforms built on this research are IMPRINT for addiction-liability screening, TOPOLOGIX for drug-protein interaction using persistent homology and bipartite simplicial complexes, and GATE for BCI neural-stimulation safety evaluation. The provisional patent on the CCT core architecture is filed for Q3 2026. Endorsements from Kent Berridge at Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU provide external validation of the theoretical framework. The next research phase requires a production database to host simulation results for replication by these collaborators and to support multi-user clinical trial simulation. Neon credits would fund this infrastructure, enabling the transition from independent research to a platform that can support collaborative validation and eventual clinical translation.

CHECKLIST

- [ ] Completed application form on Neon Startup Program website
- [ ] Motivation letter (this document)
- [ ] Short essay on technical fit (this document)
- [ ] Short essay on impact and milestones (this document)
- [ ] Research statement (this document)
- [ ] Link to ORCID profile: 0009-0001-9272-6735
- [ ] Link to GitHub profile: github.com/AmunRaPtah
- [ ] Link to ZYCO website: zyco.org
- [ ] Links to three preprints on OSF and Zenodo
- [ ] Proof of provisional patent filing (Q3 2026)
- [ ] Endorsement letters or contact information for Kent Berridge, Samuel Gershman, Nathaniel Daw, Marcelo Mattar
- [ ] CV or resume with employment history and education
- [ ] Copy of B.Pharm degree certificate and transcript
- [ ] PCN pharmacist license copy

EDITOR NOTES

- Eligibility risk: The program requires the venture to be under five years old, under fifty employees, and under five million dollars in funding. The venture has zero funding and one employee, which meets these criteria, but the applicant should confirm that independent research without a registered company entity qualifies as a startup. If a registered company is required, the applicant may need to register a Nigerian business entity before applying.
- Fact verification: The applicant states that IMPRINT currently handles 10,000 rows per query with 2.3 second latency. This number should be verified by running a benchmark on the local PostgreSQL instance and documenting the exact query and hardware specifications.
- Gap: The application does not specify the exact number of beta testers or the timeline for the first deployment. The applicant should insert a specific target number and date, for example "fifty beta testers from the University of Ibadan pharmacology department by March 2027."
- Gap: The applicant should confirm whether the Neon program requires a demo video or a pitch deck. The checklist above assumes only written materials, but the program website should be checked for additional submission requirements.
- Gap: The applicant's current employment as National Product Manager at Synthcare may raise questions about time commitment to the startup. The applicant should prepare a brief statement about how the startup work is conducted outside of work hours or with employer permission.

AI Draft — Neon Startup Program