Product Lead – Clinical Trial AI Platform (eClinical)

About Massive Bio

Every cancer patient deserves access to treatment options. Massive Bio is an AI-powered precision medicine platform transforming how cancer patients discover and access clinical trials by eliminating the barriers of geography, financial constraints, and information asymmetry that have historically limited enrollment.

Founded in 2015 and headquartered in US, Massive Bio is scaling its impact globally by powering operations across multiple countries and bringing innovative cancer treatment options to a rapidly growing and diverse population of patients. Through our proprietary AI platform, we connect individuals to clinical trials worldwide and partner with leading pharmaceutical companies, contract research organizations (CROs), and healthcare systems to accelerate drug development and expand equitable access to cutting-edge therapies.

Product Lead, SentinelAgent™ Platform

You decide what gets built, in what order, and whether a regulated sponsor can validate it. SentinelAgent™ has an approved product requirements document (PRD), a working product that runs the pipeline end to end, and a first sponsor pilot in shadow mode: running alongside the site's existing process without writing to the live database. Your job is to take it from early pilots to a validated, multi-sponsor platform without breaking a single guardrail.

Reports to: Chief Medical AI Officer

The role

You report to the Chief Medical AI Officer, with a dotted line to the CEO for commercial priorities. Our product and engineering team is in Turkey; you pair with the engineering lead there and work with our engineers, QA, design and data/ML. No direct reports to start. If you are in the US, plan to start your day by 7 a.m. Eastern: Turkey is UTC+3 year-round, so this is an East Coast seat.

You groom the backlog, write the stories, run the sprint demos and sit in the engineering reviews yourself. Four signatures gate a release. Done means your acceptance criteria pass. Validated means QA signs. Accepted means site users pass it in user acceptance testing, which our Director of Business Development & Client Delivery facilitates. Go/no-go is the Chief Medical AI Officer's call.

Twelve months out: the validation package complete, two more EDC connectors certified, and sponsors running live out of shadow mode.

What you'll do

• Own the PRD as a living document. Personas, MoSCoW-prioritized user stories with Given/When/Then acceptance criteria, functional and non-functional requirements, and full traceability: every threshold in the product traces to a requirement ID.

• Run the build with the team in Turkey. Groom the backlog in their working day, write stories an engineer can build without a meeting, demo every two weeks, and keep a release checklist nobody skips.

• Scope across the platform's modules (protocol ingestion, EMR monitoring, the EDC bridge, the ePRO connector, the closure intelligence dashboard, the retention engine, the audit ledger and the orchestrator), turning sponsor and site asks into requirements engineering can build against, and saying no where a guardrail forbids it. The guardrails: no data enters a sponsor's database without a clinician's approval, and there is no autonomous path in any configuration; the audit ledger is append-only and hash-chained; patient data never leaves its home region; model versions, including the underlying base model, are frozen per study and upgraded only between studies after revalidation; pilots run shadow-first.

• Build in the contract and billing requirements, not only the clinical ones. Logging of AI processes, prompts and outputs; no sponsor confidential information used to train or improve any model; in-region data residency for the US, EU and Japan; billing-state tracking that matches what each statement of work says.

• Make the close-out worklist the product. Required follow-up, data and queries, safety reconciliation, product accountability and essential records; named owners and deadlines; configured escalation; reviewer verification; and the audit trail behind every completed item.

• Drive the roadmap through its gates: protocol import, additional EMR and EDC connectors, ePRO and retention workflows, Japan localization, and the security and privacy attestations each gate requires.

• Own the product side of validation and integrations. User requirements, requirement-level risk assessment and traceability to test under GAMP 5 (second edition) risk-based computerized-system validation, aligned to FDA's 2024 guidance on electronic systems, records and signatures; 21 CFR Part 11 controls (audit trail, electronic signature, access control). An independent QA owner approves the validation plan, protocols and release. You define connector requirements and acceptance criteria for Medidata Rave, Oracle Clinical One and Veeva Vault EDC, ePRO platforms, EMR access through Patient Connect (our own patient app on Reticulum Nexus) and SMART on FHIR, and the Epic Vendor Services and Oracle Health developer-program tracks; engineering owns certification and listing, and Business Development owns the vendor relationship.

• Carry the regulatory-facing product work with Regulatory: regulatory-strategy materials, including the analysis that supports the product's positioning as clinician-in-the-loop decision support; model performance documentation kept inspection-ready; answers to sponsor and partner AI-governance questionnaires (base models, fine-tuning, training-data provenance, evaluation and monitoring, logs, support).

• Instrument what the PRD promises to measure (clinician correction rate, coordinator time reduction, ePRO completion, closure acceleration, concordance with monitor-verified source) and ship the readouts from the product.

What you bring

Must have

• 6+ years of product management in regulated health software, in eClinical (EDC, CTMS, eSource, eCOA), clinical AI or EHR integration, with releases shipped into GxP / 21 CFR Part 11 validated environments.

• You have owned a PRD, written acceptance criteria engineers built against, and shipped through release gates with a validation package behind you.

• Integration literacy: HL7 FHIR, CDISC ODM-XML and CDASH, EDC and ePRO APIs. You can read an API spec and a data-mapping document and find the gap.

• Working knowledge of ICH E6(R3), 21 CFR Part 11, EU Annex 11 and the EMA guideline on computerised systems and electronic data in clinical trials, GAMP 5, and privacy and cross-border transfer rules (HIPAA, GDPR, APPI) alongside sponsor data-residency commitments.

• AI/ML product experience under a locked-model regime: you have run a precision/recall or concordance evaluation and defended it to a reviewer.

• You have run product with a distributed engineering team across time zones. Professional working proficiency in English; all sponsor-facing documentation is in English.

Nice to have

• Oncology trial data: RECIST, Schedule of Assessments, long-term and survival follow-up.

• Japan: SS-MIX2 hospital systems, PMDA engagement, APPI.

• SOC 2 or HITRUST program experience; immutable or hash-chained audit logging.

• LLM-based clinical extraction and its evaluation.

• You have been the first product hire on a line before.

If you have shipped validated eClinical product and run a remote team but have not touched every standard on the list, apply anyway.

About SentinelAgent™

Massive Bio is industrializing the oncology action layer: taking a cancer patient from a clinician-confirmed trial option to an enrolled, monitored participant, and now to the follow-up, data, safety and records work that lets a site close on time. We are headquartered in Boca Raton, Florida, our product and engineering team is in Turkey, and we work with leading oncology sponsors in the United States, Europe and Asia.

SentinelAgent™ is our protocol-aware monitoring agent on the Reticulum Nexus platform. It tracks missing records, overdue follow-up and unresolved queries in one worklist, pre-fills the sponsor's EDC entry for an authorized clinician at the site to review and confirm under their own login before anything is submitted, and surfaces the open items delaying close-out for sponsor and site teams to act on. Every action sits on an append-only audit trail built to 21 CFR Part 11 expectations, and sponsor and site teams, never Massive Bio, authorize closure. SentinelAgent™ is contracted, priced and delivered separately from Massive Bio's trial-matching and patient-navigation services.