Clinical Logic Architect — Biomarker Decision Engine (Build From Scratch) | Women's Precision Health

Clarion is an early-stage startup building a precision health platform specifically for women in perimenopause, menopause, and post-menopause. The platform is designed to analyze 158 biomarkers across 28 categories and generate personalized supplement, nutrition, and exercise protocols — calibrated to each woman's hormonal stage, symptom profile, and current medications, with a tiered escalation framework that routes complex or high-risk cases to credentialed advisors (PharmD, RDN, exercise physiologist) for review. At the heart of the platform is our Logic Library: a structured repository of IF/THEN/ESCALATE-IF clinical decision rules. We are building it from the ground up — and this role owns the clinical reasoning logic at its core. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ THE ROLE — ARCHITECT THE CLINICAL LOGIC FROM THE GROUND UP ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ This is not a literature-pull role. We are not looking for someone to hand us PubMed abstracts. We are looking for a clinician who can design and build compound decision logic from scratch, grounded in actual human physiology — the reasoning that separates a safe, intelligent engine from a lookup table. This is a greenfield build. There is no legacy system to inherit or untangle — you will design the structure, the rule schema, the compound logic, and the validation approach from the ground up, and own its evolution from there. This is an ongoing engagement, not a one-time project. You will map out an extensive library of clinical scenarios and edge cases, stress-test the logic against them, and continuously optimize it as new evidence emerges, guidelines shift, and new patient permutations surface. We expect the rule set to get sharper and safer with every iteration — and we're looking for someone who wants to own that build-and-refinement loop over the long term. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ WHAT YOU'LL BUILD (FROM SCRATCH) ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ • Conditional reference-range logic — compound rules that adjust interpretation based on hormone therapy route (oral vs. transdermal), SERMs (e.g., tamoxifen vs. raloxifene, which are NOT interchangeable), oral contraceptives, assay method, and physiologic state, so each value is read in context rather than against a fixed cutoff. • Drug–nutrient and supplement–medication interaction logic, mapped to mechanism (e.g., P-gp/CYP modulation, serotonergic load, additive hypoglycemic or bleeding risk, clotting-pathway effects) — not simply flagged from a database. • Cross-marker pattern logic — compound presentations where two or three markers together tell a different story than any one alone, with explicit intervention sequencing (which deficiency to correct first, and why). • Extensive scenario mapping & continuous optimization — build out a broad library of clinical scenarios and edge cases (medication combinations, comorbidities, hormone-therapy permutations, surgical vs. natural menopause, renal/hepatic impairment, pregnancy-possible perimenopause, and more), stress-test the logic against them, and iteratively refine and optimize it over time as guidelines evolve and new cases surface. • Tiered severity and escalation logic — grounding tier thresholds and hard-stop safety rules in established clinical risk-stratification methodology. • Evidence grading — every rule graded (GRADE / Oxford CEBM), with a defined minimum evidence threshold governing whether a recommendation may auto-issue or must route to a clinician. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ THE LEVEL WE'RE HIRING FOR ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ If this way of thinking is second nature to you — if you can look at a biomarker panel and immediately account for how a patient's medications, hormone therapy, and clinical context change what each value means — this role is for you. If that framing is new, this likely isn't the right fit. ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ WHO WE'RE LOOKING FOR ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ • A clinical/medical background with real physiological depth — physician (MBBS/MD, MRCP or equivalent), PharmD, or a clinician-scientist with equivalent command of endocrinology and pharmacology. • Direct experience building or refining a biomarker-driven automated nutrition/supplement/decision engine is strongly preferred (prior work on a platform of this type is a major plus). • Fluency in evidence-grading frameworks (GRADE, Oxford CEBM) and drug-interaction pharmacology (Lexicomp, Micromedex, Natural Medicines Comprehensive Database). • Demonstrated ability to build compound, validated clinical logic from scratch — and to maintain and optimize it systematically over time, not just summarize literature. • Women's health, menopause medicine, or reproductive endocrinology experience is a significant advantage. • Comfortable working in the ambiguity of an early-stage startup — self-directed, and able to turn an open problem into structured, defensible logic.