User Personas

Document ID: PRD-007
File Path: docs/01-product/personas.md
Version: 1.0.1
Status: Draft
Owner: Product Team
Last Updated: 2026-07-07

1. Purpose

This document defines the user personas for the Wovyr AI Platform — the archetypal users whose goals and pain points drive product decisions. It expands the target-user list in the PRD §8.

2. Persona Summary

Persona	Primary need	Key surfaces
App Developer	Ship an AI feature fast	CLI, SDK, Agent Studio
AI Engineer	Build reliable agents/workflows	Agent Framework, Workflow Builder
Platform/Ops Engineer	Operate it at scale, safely	Deployment, Observability
Security/Compliance	Govern access and data	Security, Audit
Plugin Developer	Extend & distribute capabilities	Plugin SDK, Marketplace
Engineering Leader	Adopt for the org	Vision, Roadmap
Researcher	Experiment with agent designs	Agent Studio, Examples

3. Personas

3.1 App Developer — “Dana”

Context: Full-stack developer adding AI to a product; not an ML specialist.
Goals: Add a working agent/feature in days; avoid wiring many services.
Pains: Fragmented tooling, vendor lock-in, prompt-only “frameworks”.
How Wovyr helps: One platform; hello agent to production via CLI/SDK; provider independence via the LLM Gateway.
Success: First agent shipped quickly; swap models without code changes.

3.2 AI Engineer — “Amir”

Context: Builds non-trivial agents and multi-step workflows.
Goals: Grounded, reliable, debuggable agent behavior.
Pains: Non-determinism, opaque failures, brittle RAG, no durability.
How Wovyr helps: Memory Engine for RAG, durable workflows, trace/step inspection in Agent Studio.
Success: Agents that are testable (workflow tests) and observable.

3.3 Platform / Ops Engineer — “Priya”

Context: Runs the platform for many teams.
Goals: Scale, reliability, cost control, safe operations.
Pains: Weak observability, hard scaling, runaway cost.
How Wovyr helps: Kubernetes/Helm, autoscaling tool workers, SLOs/alerts (observability), quotas.
Success: Meets SLOs; cost visible and bounded per tenant.

3.4 Security / Compliance — “Sam”

Context: Owns security posture and audits.
Goals: Least privilege, isolation, auditability, compliance.
Pains: Untrusted AI tools, secret sprawl, missing audit trails.
How Wovyr helps: Sandboxed tools/plugins, RBAC/ABAC, secret vault, tamper-evident audit.
Success: Passes security testing; zero cross-tenant leakage.

3.5 Plugin Developer — “Leo”

Context: Builds integrations/tools, possibly to sell.
Goals: Package capabilities once; distribute safely.
Pains: No standard extension model; unsafe third-party code.
How Wovyr helps: Plugin SDK, signing + distribution, marketplace.
Success: A signed plugin published and installed by others (see VPN example).

3.6 Engineering Leader — “Morgan”

Context: Decides whether the org adopts Wovyr.
Goals: Productivity, control, no lock-in, sustainable roadmap.
Pains: Build-vs-buy risk, fragmented stack, governance gaps.
How Wovyr helps: Vision/business goals, open-source + provider-neutral, clear roadmap.
Success: Teams self-serve under central governance.

3.7 Researcher — “Riya”

Context: Explores agent/coordination designs.
Goals: Rapid experimentation and comparison.
Pains: Rebuilding infra per experiment; no eval harness.
How Wovyr helps: Reusable runtime, Agent Studio
- planned evaluation, multi-agent coordination.
Success: Compare agent versions on quality/cost/latency easily.

4. Anti-Personas (Not Targeted)

Per PRD scope: consumers seeking a chat app, teams wanting a low-code website builder, or users needing foundation-model training are not the target.

6. Revision History

Version	Date	Description
1.0.1	2026-07-07	Renumbered from PRD-002 to PRD-007 — that ID collided with `prd-future.md`, which was independently assigned PRD-002 later without checking the sequence. Found during a project-wide doc review; no content changed
1.0.0	2026-06-27	Initial User Personas