MOTIVATION LETTER The Generation H HealthTech initiative at SET University is designed for ventures that have moved beyond concept into product, and that is exactly where psyche-twin stands today. The typed, multi-scale knowledge-graph architecture is built, the append-only event log is functional, and the first-person conversational interface generates responses grounded in the graph's current state. This is a working system, not a slide deck or a simulation. It fuses multiple independent evidence streams, including LLM-derived, assessment-derived, behavioral, and document-derived inputs, into a single coherent model of a user's mental state. My background is unusual for a founder, and it is precisely why this venture is technically viable. I am a licensed pharmacist with a B.Pharm from the University of Ibadan, a computational modeler who has built and calibrated ODE-based brain-circuit simulations, and a software engineer who has shipped production systems. The neurocascade engine, a receptor-to-behavior simulation with 62 passing tests and Bayesian calibration via PyMC, gives me a quantitative understanding of the neurobiological signals that psyche-twin ingests. The CCT model, a tripartite pharmacological framework for reward-memory encoding, taught me how to handle multi-stream, noisy, time-series data with rigorous Bayesian methods. psyche-twin is the applied product layer built on top of that methodological foundation. The program's selection criteria are clear: an MVP, confirmed traction, two committed representatives, and a serious intent to pursue investment and monetization. psyche-twin meets the MVP requirement today. The disagreement-as-edge design, where conflicting evidence streams become explicit graph edges rather than being averaged away, is a defensible technical differentiator in a mental health market crowded with chatbots that flatten user data into generic responses. The monetization path is direct: a subscription-based digital twin for individuals, with a licensed clinical dashboard for therapists and clinics. The Africa angle is a market advantage, not a footnote. Nigeria has fewer than 500 practicing clinical psychologists for over 200 million people, and mobile penetration is above 90 percent. A graph-based self-model that runs on a phone and does not require a human therapist for basic reflective support addresses a structural shortage, not a luxury niche. I am committed to the three-month intensive format, including offline and online engagement, and I will bring a second project representative who is prepared to sign the participation agreement and join the SET Community. The goal is to leave with a term sheet, a pilot partner, and a clear hiring plan, not to attend a workshop. Generation H's focus on Ukrainian and global HealthTech ventures matches my own operational reality: I am enrolled in the M.Sc. Digital Health program at Hasso Plattner Institute in Potsdam, Germany, starting Winter Semester 2026/27, and I currently manage national product operations at Synthcare. I build systems that work in constrained environments. That is what psyche-twin is, and that is what I will bring to SET University. RESEARCH STATEMENT psyche-twin is a typed, multi-scale knowledge-graph architecture for self-modeling in mental health. The core technical problem it solves is data fusion under disagreement. Standard mental health apps collect self-reports, passive sensor data, and questionnaire scores, then average them into a single trend line, which destroys the signal that matters most: the moments when different evidence streams contradict each other. psyche-twin treats disagreement as first-class data. Each evidence stream, whether derived from an LLM-based conversational interface, a standardized assessment, a behavioral log, or an uploaded document, writes events into a single append-only log. When two streams conflict, the system creates an explicit graph edge labeled as a disagreement, rather than smoothing it away. The user's current state is always a function of the graph's current topology, not a hidden latent variable. The architecture is already implemented. The knowledge graph is typed, meaning every node and edge has a schema, which allows for rigorous querying and validation. The append-only event log ensures auditability; no data is ever overwritten, only superseded. The conversational interface is a first-person system, meaning it speaks from the graph's current state, not from a generic therapeutic script. If the graph shows a disagreement between a user's self-reported mood and their behavioral data, the interface can surface that contradiction directly, which is a clinically meaningful intervention in itself. My technical background directly informs this design. I built neurocascade, a receptor-to-behavior brain-circuit simulation engine that couples pharmacokinetics to receptor binding to Wilson-Cowan circuit dynamics to behavioral readouts. That engine taught me how to model multi-scale systems where parameters at one level constrain dynamics at another. psyche-twin is the same problem in reverse: instead of simulating from receptors up to behavior, it infers from behavioral and self-report data down to a structured model of the user's cognitive and emotional state. The Bayesian calibration methods I used for neurocascade, specifically PyMC with DEMetropolisZ sampling, are directly applicable to the inference problems psyche-twin will face as it scales. The validation status is honest and specific. The system is functional as a software artifact; all core modules pass their test suites, and the graph architecture handles concurrent writes from multiple streams without corruption. What has not yet happened is a real-data validation gate with human users. I have not claimed product-market fit, revenue, or clinical validation, because those claims would be false. The next step, which this program directly supports, is a pilot study with 20 to 30 users over eight weeks, measuring whether the disagreement-edge visualization improves users' ability to identify their own cognitive biases compared to a standard mood-tracking baseline. That pilot is designed, pre-registered, and ready to launch. The market context is urgent. The global digital mental health market is projected to exceed USD 20 billion by 2027, yet most products are shallow. They collect data, they do not model the person. psyche-twin's multi-scale, disagreement-preserving architecture is a genuine technical advance, not a feature tweak. It is built by someone who understands both the neurobiology and the software engineering, and it is ready for the intensive validation and go-to-market push that Generation H provides. SHORT ANSWER ESSAY: TEAM AND COMMITMENT The program requires at least two project representatives committed to intensive offline and online engagement for three months. I am the primary representative. I am a licensed pharmacist, a computational researcher with peer-reviewed work under review at three journals, and the current National Product Manager at Synthcare, where I oversee product operations across Nigeria. I will bring a second representative, a software engineer with whom I have collaborated on the psyche-twin codebase for the past six months. That engineer has deep experience in graph databases and real-time systems, and is prepared to commit to the full program schedule. We have already agreed on a division of labor: I own the clinical model, the evidence-stream design, and the regulatory strategy; my collaborator owns the graph infrastructure, the conversational interface, and the deployment pipeline. Both of us are prepared to sign the participation agreement and join the SET Community. We are attending to build, test, and raise, not to observe. SHORT ANSWER ESSAY: COMMERCIAL INTENT The intent to develop psyche-twin as a commercial venture is explicit and operational. The monetization model has two tiers. The first is a direct-to-consumer subscription for the personal self-model, priced competitively against existing mental health apps. The second is a licensed clinical dashboard for therapists and clinics, which provides a structured view of a patient's graph state between sessions, reducing no-show rates and improving treatment continuity. The revenue model is based on comparable pricing in the digital therapeutics market, not speculation. The job-creation plan is concrete: the first hire after a successful pilot will be a clinical validation lead, followed by a growth marketer focused on the Nigerian and broader African market. The investment thesis is equally specific. A seed round of USD 500,000 would fund the eight-week pilot, the clinical dashboard build, and the first six months of operations. The exit path is acquisition by a larger digital health platform or a pharmaceutical company with a mental health portfolio, both of which are active acquirers in this space. I am building a company with a research-grade core, not a research project with a business plan attached. CHECKLIST - [ ] Confirm the exact application deadline on the SET University Generation H page - [ ] Verify the program's current cohort dates and whether the three-month format is still active - [ ] Confirm whether the program accepts solo founders with an external collaborator or requires both representatives to be co-founders - [ ] Prepare a one-page pitch deck for psyche-twin, including architecture diagram and pilot design - [ ] Prepare a demo video of the psyche-twin conversational interface, under three minutes - [ ] Obtain a signed commitment letter from the second project representative - [ ] Prepare a two-page technical appendix describing the graph schema and the disagreement-edge design - [ ] Prepare a one-page budget table for the eight-week pilot study - [ ] Verify English proficiency documentation, if required (Upper-Intermediate or higher) - [ ] Confirm whether any participation fee or equity requirement applies to accepted startups - [ ] Prepare a list of three references, including one academic and one industry contact - [ ] Draft and sign the participation agreement, if the program provides one - [ ] Prepare a short statement on how psyche-twin will contribute to the SET Community beyond the program duration EDITOR NOTES - Eligibility risk: the program may require a registered legal entity for the startup; psyche-twin is currently an independent project under Eniola's name. Verify whether a company registration is needed before application. - The second project representative is not named in the applicant profile. Eniola must insert the actual collaborator's name, background, and a signed commitment letter before submission. - The pilot study design is referenced as pre-registered but no registry link or DOI is provided. Eniola must supply the OSF or equivalent registration URL to substantiate this claim. - The revenue model cites comparable pricing in digital therapeutics but no specific figures are given. Eniola should insert concrete price points and a one-line market comparables table. - The program's selection criteria mention "minimum confirmed revenues or sales volumes." psyche-twin has none. The application must address this gap directly, either by reframing the pilot as a paid beta with confirmed letters of intent from clinics, or by explicitly stating that revenue is not yet applicable and leaning on the MVP and traction criteria instead.