# Diaphora.ai Brand Voice Guidelines

This document outlines the communication style, tone, and patterns for Diaphora.ai, ensuring consistency and effectiveness across all brand content.

### Communication Style
-   **Overall tone and personality:** Authoritative, precise, confident, pragmatic, technical, and solution-oriented. Diaphora.ai speaks as a knowledgeable expert and a trusted engineering partner, understanding and addressing the complex challenges of building reliable AI in production. It respects the intelligence of its technical audience, avoiding fluff and buzzwords in favor of clarity and technical depth.
-   **Key stylistic elements and patterns:**
    -   **Direct Problem-Solution Framing:** Clearly articulates a problem ("Your best skills are still just prompts.") then offers a precise solution or benefit ("Schema-validated. Same shape, every run."). Often uses "Before" vs. "After" comparisons.
    -   **Declarative and Confident Language:** Uses strong, active verbs and makes definitive statements ("That's the point.", "Diaphora makes that work unnecessary.").
    -   **Emphasis on Key Attributes:** Frequently highlights words like "reliable," "typed," "governed," "auditable," "production-ready," "structured," and "deterministic."
    -   **Technical Precision:** Integrates industry-specific terminology and concepts naturally, assuming a baseline technical understanding.
    -   **Contrasting for Impact:** Often contrasts its approach with common pitfalls or less effective alternatives ("No drag-and-drop canvas. That's the point.").
    -   **Structured Information:** Utilizes bullet points, bolding, and clear headings to break down complex information.
-   **Vocabulary preferences and word choices:** Highly technical and domain-specific. Favors terms related to AI/ML engineering, backend services, data governance, and software development. Examples include: *blueprint, FML, MCP server, LLM agent, schema-validated, RBAC, ABAC, DLP, PII, SSO, SAML, OpenAPI, inference, session, transformer, context, token limits, hallucination, deterministic, auditable.* Avoids vague marketing jargon.

### Content Patterns
-   **Common themes and topics:** The core themes revolve around graduating AI "skills" into robust, production-grade backend services. This includes reliability, governance, structured and predictable output, scalability, integration with existing systems, and a developer-centric experience for engineers building AI automations.
-   **Structural approaches to content:**
    -   **Problem-first approach:** Starts by articulating a common pain point or inefficiency, then positions Diaphora.ai as the definitive solution.
    -   **Feature deep-dives:** Explains features with technical detail, often including code snippets (e.g., CLI commands, FML examples) to illustrate functionality.
    -   **Comparative analysis:** Directly compares its approach to alternative methods or tools, highlighting its advantages.
    -   **Clear segmentation:** Content is organized with distinct headings and subheadings for easy navigation and comprehension of complex topics.
-   **Call-to-action styles and patterns:** CTAs are direct, clear, and action-oriented. They are strategically placed after a problem statement or a compelling feature explanation. Examples include: "Book a demo," "Start free," "Convert a skill file," "View the open-source runtime on GitHub." Often provides alternative or supplementary CTAs like "[Or book a demo →]" or "[See how this compares... →]".

### Audience Interaction
-   **How the brand addresses its audience:** Directly, using "you" and "your," assuming the audience is an engineer or technical specialist grappling with the challenges Diaphora.ai solves. It speaks *to* the audience as a peer who understands their technical needs and frustrations.
-   **Level of formality and relationship style:** Professional and expert-to-expert. The relationship is built on mutual technical understanding and trust in competence, rather than casual familiarity. It positions Diaphora.ai as a reliable partner in solving complex engineering problems.
-   **Engagement and conversation patterns:** Engagement is driven by articulating and validating the audience's technical pain points, then providing precise, actionable solutions. The conversation is focused on technical merit, functionality, and tangible benefits for production environments. It's about empowering engineers to build better, more reliable AI systems.

### Guidelines & Examples
-   **Do's for brand communication:**
    -   **Be technically precise:** Use correct terminology and explain complex concepts clearly.
    -   **Focus on reliability and production-readiness:** Emphasize how Diaphora.ai ensures consistent, governed, and scalable AI automations.
    -   **Address engineer pain points directly:** Show empathy for the challenges of building robust AI.
    -   **Provide concrete examples:** Use code snippets, CLI commands, or FML examples to illustrate functionality.
    -   **Be confident and authoritative:** Position Diaphora.ai as the definitive solution for its niche.
    -   **Highlight governance, structure, and auditability.**
-   **Don'ts for brand communication:**
    -   **Use vague or generic marketing jargon:** Avoid terms that lack specific technical meaning.
    -   **Be overly casual or informal:** Maintain a professional, expert-level tone.
    -   **Oversimplify technical concepts to the point of inaccuracy.**
    -   **Make unsubstantiated claims:** Back up benefits with clear functional descriptions.
    -   **Avoid technical details:** Our audience expects depth.
-   **Example phrases and expressions that are "on-brand":**
    -   "Turn AI-powered automations into typed, reliable backend services."
    -   "Graduate your skills into services."
    -   "Schema-validated. Same shape, every run."
    -   "The OSS runtime for AI-powered backend services."
    -   "There's no drag-and-drop canvas. That's the point."
    -   "Ship a typed, deterministic blueprint you can review, diff, and version like real code."
    -   "Governed automatically — RBAC, ABAC, and DLP enforced on every run."
-   **Content types and formats the brand uses:** Website copy, product descriptions, technical documentation (guides, API reference, SDKs), blog posts (likely deep-dives into specific use cases, technical explanations, or comparisons), GitHub READMEs, and CLI output examples.