Skip to main content

CODITECT Product Development Research Prompts v2

Structured Prompts for Deep-Dive Analysis​

This document provides categorized, well-structured prompts to extend the Palantir strategic analysis into actionable CODITECT product development directions.

Category 1: Ontology & Data Architecture​

Prompt 1.1: Healthcare Domain Ontology Design​

CONTEXT:
Palantir's Ontology is central to their competitive advantage - a semantic layer that provides business context for AI operations. CODITECT needs a compliance-native ontology for healthcare.

TASK:
Design a healthcare compliance ontology for CODITECT that includes:
1. Core object types (Patient, Encounter, Claim, Provider, ClinicalDocument, ComplianceCheck)
2. Relationship types between objects
3. Property definitions with compliance metadata (PHI indicators, retention rules)
4. Action types that represent compliant state transitions
5. Audit trail integration patterns

OUTPUT FORMAT:
Return a TypeScript interface definition with:
- Object type schemas with JSDoc compliance annotations
- Link type definitions with cardinality
- Action type signatures with pre/post conditions
- Example instantiations for a prior authorization workflow

CONSTRAINTS:
- All PHI fields must have @hipaa annotation
- Audit requirements from FDA 21 CFR Part 11 must be addressable
- Design must support row-level security inheritance

Prompt 1.2: Financial Services Compliance Objects​

CONTEXT:
Financial services compliance requires tracking transactions, regulatory filings, and audit trails across SOX, SOC 2, and international standards like LGPD.

TASK:
Create a financial compliance ontology extension for CODITECT covering:
1. Transaction objects with regulatory classification
2. Control testing objects for SOX compliance
3. Evidence collection objects for SOC 2 audits
4. Cross-border data flow tracking for LGPD/GDPR
5. Segregation of duties modeling

OUTPUT FORMAT:
JSON Schema definitions with:
- Object types and their compliance categorization
- Required vs optional fields with regulatory mapping
- Relationship definitions for control-to-evidence linking
- Example: Complete SOX control testing workflow

REFERENCE:
Use Palantir's ontology patterns from their FY 2024 10-K documentation as a structural guide, but specialize for mid-market financial institutions.

Prompt 1.3: Cross-Domain Compliance Unification​

CONTEXT:
Many organizations operate across healthcare AND financial services (e.g., health insurance, HSA providers). CODITECT needs a unified compliance model.

TASK:
Design a meta-ontology that:
1. Unifies healthcare (HIPAA) and financial (SOX) compliance requirements
2. Identifies shared patterns (audit trails, access controls, data retention)
3. Models conflict resolution when requirements diverge
4. Supports multi-jurisdictional compliance (US + Brazil LGPD)

OUTPUT FORMAT:
- UML class diagram (PlantUML syntax)
- Conflict resolution decision tree
- JSON mapping between healthcare and financial object types
- Implementation priority matrix (high impact, low complexity first)

Category 2: Agent Architecture & AgentOps​

Prompt 2.1: Guardrail-Constrained Agent Framework​

CONTEXT:
Palantir's AIP binds LLM actions to role-based access, approvals, and audit trails. CODITECT agents must operate within strict compliance guardrails.

TASK:
Design a guardrail framework for CODITECT compliance agents:
1. Define guardrail types (access control, action limits, approval gates, audit requirements)
2. Specify how guardrails compose (AND/OR/sequential)
3. Model human-in-loop escalation patterns
4. Design fallback behaviors when guardrails block actions
5. Implement audit trail generation for all guardrail evaluations

OUTPUT FORMAT:
Python code with:
- Abstract Guardrail class with validation interface
- Concrete implementations (AccessGuardrail, ApprovalGuardrail, AuditGuardrail)
- GuardrailChain composition class
- Integration example: Prior authorization agent with PHI guardrails

CONSTRAINTS:
- Must support real-time evaluation (<100ms p95)
- Must produce structured audit events
- Must handle guardrail failures gracefully (never expose raw errors)

Prompt 2.2: Multi-Agent Orchestration for Compliance Workflows​

CONTEXT:
Complex compliance workflows (e.g., FDA pre-submission, SOX control testing) require multiple specialized agents working together.

TASK:
Design a multi-agent orchestration pattern for CODITECT:
1. Define agent roles (Researcher, Validator, Documenter, Reviewer)
2. Specify inter-agent communication protocol
3. Model shared state management
4. Design checkpoint/recovery mechanisms
5. Implement compliance-aware task routing

OUTPUT FORMAT:
- Agent role specifications (JSON schema)
- Sequence diagram for a complete FDA 510(k) pre-submission workflow
- Python orchestrator implementation using async patterns
- Error handling and human escalation flow

REFERENCE:
Use Anthropic's orchestrator-workers and evaluator-optimizer patterns as the foundation.

Prompt 2.3: AgentOps Monitoring & Evaluation​

CONTEXT:
FDE roles are evolving into Agent Deployment Engineers who must monitor, evaluate, and maintain agent networks in production.

TASK:
Design an AgentOps monitoring system for CODITECT:
1. Define key metrics (accuracy, safety, latency, cost, compliance rate)
2. Design evaluation pipelines (offline test suites, online A/B testing, red-teaming)
3. Specify alerting thresholds and escalation paths
4. Model agent versioning and rollback procedures
5. Create dashboard specifications

OUTPUT FORMAT:
- Metrics taxonomy with collection methods
- Evaluation pipeline architecture (Mermaid diagram)
- Alert rule specifications (Prometheus-style)
- Dashboard wireframe with key visualizations
- Python implementation for core evaluation functions

CONSTRAINTS:
- Must support regulated industry audit requirements
- Must enable "explain this decision" for any agent action
- Must integrate with existing observability stacks (Datadog, Grafana)

Category 3: Go-to-Market & Bootcamp Model​

Prompt 3.1: 2-Day Compliance Bootcamp Design​

CONTEXT:
Palantir's 5-day AIP bootcamp achieves ~60% conversion. CODITECT needs a compressed 2-day model for mid-market compliance automation.

TASK:
Design a complete 2-day bootcamp curriculum:
1. Hour-by-hour schedule with activities and deliverables
2. Pre-work requirements to compress timeline
3. Technical setup automation (infrastructure, data integration)
4. Demo scenarios for healthcare and financial services
5. ROI calculation methodology and presentation template

OUTPUT FORMAT:
- Detailed agenda (JSON with time blocks, activities, owners, deliverables)
- Pre-bootcamp checklist
- Technical setup scripts (bash/Python)
- Demo script for prior authorization automation
- ROI calculator spreadsheet specification
- Post-bootcamp follow-up sequence

SUCCESS CRITERIA:
- Customer can see 2x ROI path within 20 days
- Technical integration complete by end of Day 1
- At least one production-ready agent deployed by end of Day 2
- Clear next steps and pricing discussion initiated

Prompt 3.2: Mid-Market Pricing Strategy​

CONTEXT:
Palantir targets $5M+ enterprise deals. CODITECT targets $50K-$500K mid-market with compliance premium.

TASK:
Develop a pricing strategy for CODITECT:
1. Analyze Palantir, Snowflake, and Databricks pricing models
2. Define CODITECT pricing tiers (Starter, Growth, Enterprise)
3. Model unit economics (CAC, LTV, payback period)
4. Calculate compliance premium justification
5. Design consumption vs subscription hybrid model

OUTPUT FORMAT:
- Pricing tier table with features and limits
- Unit economics model (spreadsheet specification)
- Competitive positioning matrix
- Value-based pricing calculator for sales team
- Discount authority guidelines

CONSTRAINTS:
- Target CAC payback <3 months
- Target LTV:CAC >3:1
- Must support annual and multi-year contracts
- Compliance modules must have clear pricing attribution

Prompt 3.3: Reference Customer Development​

CONTEXT:
CODITECT needs 3 reference customers within 6 months to validate the model and support sales.

TASK:
Design a reference customer development program:
1. Ideal customer profile (ICP) for first 3 customers
2. Success criteria and milestone definitions
3. Case study template and production process
4. Reference call program structure
5. Co-marketing opportunity identification

OUTPUT FORMAT:
- ICP scorecard with weighted attributes
- Customer success playbook (90-day plan)
- Case study template (problem, solution, results, quote)
- Reference call script and guidelines
- Co-marketing proposal template

TARGET OUTCOMES:
- Each reference customer achieves documented ROI >2x
- Public case study published within 90 days of go-live
- Reference willing to do sales calls and conference speaking

Category 4: Compliance Framework Deep Dives​

Prompt 4.1: FDA 21 CFR Part 11 Implementation Guide​

CONTEXT:
FDA 21 CFR Part 11 governs electronic records and signatures in pharmaceutical and medical device industries. CODITECT must be Part 11-compliant by design.

TASK:
Create a comprehensive Part 11 implementation guide for CODITECT:
1. Map Part 11 requirements to platform capabilities
2. Design electronic signature workflows
3. Specify audit trail requirements and implementation
4. Document access control patterns
5. Create validation documentation templates (IQ/OQ/PQ)

OUTPUT FORMAT:
- Requirements traceability matrix (Excel specification)
- Electronic signature implementation guide
- Audit trail schema with example records
- Access control matrix template
- Validation protocol templates
- Gap analysis methodology

REGULATORY REFERENCE:
FDA 21 CFR Part 11, FDA Guidance for Industry (Part 11 Scope and Application)

Prompt 4.2: HIPAA Technical Safeguards Automation​

CONTEXT:
HIPAA technical safeguards require access controls, audit controls, integrity controls, and transmission security. CODITECT agents must automate compliance verification.

TASK:
Design automated HIPAA technical safeguard verification:
1. Define checkpoints for each safeguard category
2. Create automated evidence collection agents
3. Design continuous compliance monitoring
4. Build remediation suggestion engine
5. Generate audit-ready documentation

OUTPUT FORMAT:
- Safeguard checklist with automation status
- Agent definitions for each verification task
- Monitoring dashboard specification
- Remediation playbook templates
- Sample audit report

HIPAA REFERENCE:
45 CFR 164.312 (Technical safeguards)

Prompt 4.3: EU AI Act Compliance Architecture​

CONTEXT:
The EU AI Act enters enforcement in August 2026. High-risk AI systems (including healthcare and financial AI) require extensive documentation, testing, and human oversight.

TASK:
Design CODITECT's EU AI Act compliance architecture:
1. Risk classification framework for customer AI systems
2. Required documentation templates (technical files, conformity declarations)
3. Human oversight integration patterns
4. Bias detection and mitigation workflows
5. Incident reporting automation

OUTPUT FORMAT:
- Risk classification decision tree
- Documentation templates (markdown)
- Human oversight workflow diagrams
- Bias detection pipeline specification
- Incident report schema and routing logic

REGULATORY REFERENCE:
EU AI Act (Regulation 2024/1689), with focus on Articles 9-15 (high-risk requirements)

Category 5: Technical Architecture​

Prompt 5.1: Tiered LLM Routing Implementation​

CONTEXT:
Palantir uses model routing to optimize cost/quality tradeoffs. CODITECT should route: Opus for regulatory/security, Sonnet for standard, Haiku for simple tasks.

TASK:
Design a tiered LLM routing system:
1. Define task classification taxonomy
2. Build routing decision function
3. Implement fallback and retry logic
4. Design cost tracking and reporting
5. Create A/B testing framework for routing optimization

OUTPUT FORMAT:
- Task taxonomy with routing rules (JSON)
- Python routing implementation with type hints
- Cost tracking schema and dashboard specification
- A/B testing framework design
- Projected cost savings analysis (40-70% target)

CONSTRAINTS:
- Routing decision must be <10ms
- Must support dynamic model availability
- Must maintain audit trail of routing decisions
- Must enable manual override for specific tasks

Prompt 5.2: Compliance-Native Audit Trail Architecture​

CONTEXT:
Regulated industries require immutable, complete audit trails. CODITECT's audit system must be a first-class architectural component.

TASK:
Design a compliance-native audit trail system:
1. Define audit event schema (who, what, when, where, why)
2. Implement hash chain for immutability
3. Design retention policies by regulation
4. Build query interface for investigations
5. Create export formats for auditors

OUTPUT FORMAT:
- Audit event schema (JSON Schema)
- Hash chain implementation (Python)
- Retention policy configuration system
- Query API specification (OpenAPI)
- Export templates (CSV, PDF, XBRL for financial)

CONSTRAINTS:
- Must support 7-year retention (FDA) and 10-year (SOX)
- Must be tamper-evident (detect unauthorized modifications)
- Must support point-in-time reconstruction
- Must integrate with external SIEM systems

Prompt 5.3: Multi-Deployment Architecture (Cloud/On-Prem/Hybrid)​

CONTEXT:
Palantir's Apollo enables deployment across cloud, on-prem, and air-gapped environments. CODITECT must support customer deployment flexibility.

TASK:
Design a multi-deployment architecture for CODITECT:
1. Define deployment topologies (SaaS, customer cloud, on-prem)
2. Design data residency controls
3. Build configuration management system
4. Create deployment automation (IaC)
5. Implement health monitoring across topologies

OUTPUT FORMAT:
- Deployment topology diagrams (C4 model)
- Data residency configuration schema
- Terraform/Pulumi modules for each topology
- Health check and monitoring specification
- Upgrade and rollback procedures

CONSTRAINTS:
- Must support Kubernetes and bare metal
- Must enable data sovereignty compliance
- Must maintain feature parity across topologies
- Must support air-gapped deployment for sensitive environments

Category 6: Competitive Intelligence​

Prompt 6.1: Palantir AIP Feature Teardown​

CONTEXT:
Palantir AIP is the primary competitive threat. Deep understanding of its capabilities informs CODITECT differentiation.

TASK:
Conduct a comprehensive AIP feature teardown:
1. Catalog all publicly documented AIP capabilities
2. Analyze AIP bootcamp case studies for feature patterns
3. Identify integration points with Foundry/Gotham
4. Map features to customer outcomes
5. Identify gaps and CODITECT differentiation opportunities

OUTPUT FORMAT:
- Feature catalog (hierarchical, with descriptions)
- Case study analysis matrix
- Architecture diagram of AIP stack
- Feature-to-outcome mapping table
- Competitive differentiation matrix

SOURCES:
Palantir documentation, AIPCon presentations, customer case studies, investor materials

Prompt 6.2: Mid-Market Compliance Automation Landscape​

CONTEXT:
CODITECT targets mid-market compliance automation. Understanding the full competitive landscape is critical.

TASK:
Map the mid-market compliance automation competitive landscape:
1. Identify direct competitors (compliance automation startups)
2. Analyze adjacent players (GRC platforms, audit software)
3. Evaluate horizontal AI platforms entering compliance
4. Assess "build vs buy" alternatives
5. Identify partnership opportunities

OUTPUT FORMAT:
- Competitive landscape map (visual)
- Competitor profiles (5-10 key players)
- Feature comparison matrix
- Pricing comparison table
- SWOT analysis for CODITECT

FOCUS AREAS:
Healthcare (HIPAA, FDA), Financial (SOX, SOC 2), Cross-industry (EU AI Act)

Prompt 6.3: Palantir Healthcare Expansion Analysis​

CONTEXT:
Palantir's NHS success suggests healthcare expansion. CODITECT must anticipate and counter this threat.

TASK:
Analyze Palantir's healthcare strategy and CODITECT's defensive position:
1. Document Palantir's healthcare wins (NHS, US health agencies)
2. Identify Palantir's healthcare-specific capabilities
3. Assess likelihood of mid-market healthcare push
4. Design CODITECT's defensive moat
5. Identify acquisition/partnership targets to accelerate

OUTPUT FORMAT:
- Palantir healthcare timeline and wins
- Healthcare capability assessment
- Threat probability analysis
- Defensive strategy recommendations
- M&A/partnership target list with rationale

TIMEFRAME:
Focus on 2026-2027 strategic window before potential Palantir mid-market expansion

Usage Guidelines​

Prompt Selection​

  1. Immediate priorities (Phase 1):

    • Prompt 1.1 (Healthcare Ontology)
    • Prompt 2.1 (Guardrail Framework)
    • Prompt 3.1 (Bootcamp Design)

  2. Validation priorities (Phase 2):

    • Prompt 2.3 (AgentOps Monitoring)
    • Prompt 4.1 (FDA 21 CFR Part 11)
    • Prompt 3.3 (Reference Customers)

  3. Scale priorities (Phase 3):

    • Prompt 1.2 (Financial Ontology)
    • Prompt 5.3 (Multi-Deployment)
    • Prompt 4.3 (EU AI Act)

Execution Pattern​

for prompt in prioritized_prompts:
    # 1. Execute prompt with Claude
    response = claude.complete(prompt)
    
    # 2. Review for compliance alignment
    compliance_check = validate_regulatory_references(response)
    
    # 3. Extract actionable artifacts
    artifacts = extract_code_schemas_diagrams(response)
    
    # 4. Integrate into CODITECT backlog
    create_jira_tickets(artifacts)
    
    # 5. Schedule follow-up prompts
    queue_dependent_prompts(prompt.children)

Document Version: 2.0 | Date: February 2026 Based on: Palantir Q4 2025 Analysis, FDE→AgentOps Evolution, Competitive KPIs