Phase 01
Scoped discovery
Two week engagement to map the problem, enumerate failure modes, and publish a scoped statement of work. Fixed fee. Fully refundable against the first phase of implementation if the scope concludes unfavourably.
Engagements
Narrow scope.
Clinical
review.
I.
Line 01
Governed AI platforms.
Custom governed AI systems built on the Externally Governed Learning Systems framework (US Provisional Patent 63/975,551). Delivered as code with complete governance declarations and audit artifacts.
Critic actor pipelines for discovery, synthesis, and education
Pipeline architectures for content generation, protocol summarisation, literature triage, clinical education workflows, and drug discovery campaigns. Generation agents are separated from evaluative agents by contract. Quality gates are declared at named stages and enforced through bounded retry policies.
- Multi tier agent architecture with declared quality thresholds, following the DrugSynthAI reference design (five tiers, sixty seven agents in production)
- Per agent telemetry logged to a governance database for audit, regression detection, and post hoc review
- Cost aware model routing with local first inference through Ollama and cloud fallback through the MedClaw runtime where appropriate
- Full handover of code, configuration, governance declarations, and operational runbooks to the client engineering and clinical teams
What the engagement produces
- Working pipeline repository with reproducible environment and deployment scripts
- Governance declaration documenting agent responsibilities and authority boundaries
- Operational runbook covering failure modes, retry policies, and recovery procedures
- Training sessions for the receiving engineering and clinical teams
- Optional continuation under the AIDD-GOV open standard for drug discovery engagements
II.
Line 02
Clinical operations software.
Patient facing applications and physician operating systems. FxMED OS is available to selected partner practices under a technology partnership and business associate agreement where protected health information is in scope.
Physician facing operating system in production
FxMED OS is live in production with nineteen complete service layers: patient registry, scheduling, queue flow, encounter management, clinical documents, consent and permissions, attachments and media, portal policy, communications, AI scribe, tenancy, task workflow, entitlements, search, onboarding, white label, proxy access, membership, CRM, and guided setup.
- Standard API integration with existing CRM, EHR, and laboratory systems
- Role based access control and audit friendly logging at every integration point
- Notification and reminder workflows tuned to clinical adherence objectives
- HIPAA aligned deployment posture under an executed business associate agreement
- ES256 JWT authentication with legacy HS256 fallback, governance ledger (SOGDR), Watcher Protocol audit
What the firm does not do
FxMEDUS does not deliver direct patient care, does not issue clinical decisions outside of Dr. Borges's own clinical practice, and does not process protected health information outside of an executed business associate agreement with the engaging clinical entity. Software deployed in regulated contexts is scoped, documented, reviewed, and audited accordingly.
III.
Line 03
Research infrastructure.
Structured evidence synthesis, preprint triage, literature surveillance, and drug discovery campaigns for principal investigators, academic departments, and industrial partners. Methods are published, reproducible, and transferable.
Literature, evidence, and discovery infrastructure
Pipelines that ingest, screen, and synthesise biomedical literature on a continuous schedule. In drug discovery contexts, the DrugSynthAI platform supports governed campaigns covering target selection, de novo molecule design, ADMET profiling, and stage gate review.
- Search strategy design across indexed databases (PubMed, Scopus, Embase) and preprint servers (bioRxiv, medRxiv, ChemRxiv, Research Square)
- Rolling meta analysis with explicit temporal weighting and declared inclusion rules
- Structured outputs suitable for downstream dashboards, editorial workflows, grant preparation, or systematic review submission
- Governance artifacts supporting systematic review registration, PRISMA reporting, and compliant data release on Harvard Dataverse or equivalent
- Drug discovery campaigns with documented stage gates, candidate ranking, and kill switch protocols
IV.
Engagement model
How the firm works.
Every engagement begins with a scoped discovery and ends with a documented handover. Nothing is built without an explicit success criterion and a declared failure mode.
Phase 02
Supervised build
Implementation under physician led technical supervision. Weekly progress review with declared thresholds at every stage gate. All code, governance artifacts, and documentation handed over at phase close.
Phase 03
Transfer and support
Knowledge transfer to the client engineering and clinical teams. Optional retained support on a defined monthly envelope. No lock in clauses. All client data and artifacts portable at contract termination.
V.
Next step
Begin with a scoped discovery.
Identify the clinical, research, or engagement problem you intend to solve. The firm will respond within one business day with a proposed scope, timeline, and fee envelope.