An Autonomous AI Agent Writing Your Content
A production AI agent that writes research-backed blog posts and pages, reviews them against your brand voice rules, and publishes to WordPress on a fixed schedule. No prompting required.
Aria is a managed autonomous AI agent that runs a complete content production operation for your business. She takes research briefs, writes full drafts, reviews them against documented brand voice rules, generates images, and publishes finished content to WordPress — all on a fixed cron schedule, with no per-piece prompting, running on the same AI workflow infrastructure we use for all our autonomous agent deployments.
THE CHALLENGE
The Content Production Problem You Can’t Staff Around
Hire Writers
$400–500+ per fully researched, SEO-optimized article with editing rounds
✗ Quality variance between writers and assignments
✗ Onboarding takes weeks before they match your voice
✗ Deadlines slip when they juggle multiple clients
✗ Knowledge walks out when they leave
Do It Yourself
Your time
✗ You know what to say, never find time to write it
✗ Falls to the bottom of the list every week
✗ Inconsistent publishing kills SEO compound gains
✗ Context-switching between writing and running a business
Content Mills
Cheap per word
✗ Generic output that reads like every other company’s blog
✗ No brand voice. Heavy editing required to not sound like AI filler.
✗ Surface-level research, recycled talking points
✗ No understanding of your business or audience
Each of these approaches works in the right context. The common failure is consistency. Content marketing compounds over time, and compound gains require sustained output at a quality threshold. One good article per quarter does not move the needle. Twenty solid articles over six months does. That sustained cadence is what most teams struggle to maintain. According to Gartner, fewer than 5% of enterprise apps embedded task-specific agents in 2025, but that number is forecast to reach 40% by the end of 2026. The infrastructure for autonomous content production is not speculative — it is operational.
AUTONOMOUS CONTENT AGENT
Meet Aria — An Agentic AI Content Writer
Aria is an autonomous AI agent that runs a complete content production operation on a fixed schedule. She reads research briefs, loads your documented brand voice rules, writes a full draft, runs a two-pass self-review, generates images, and publishes to WordPress. Every step fires on a cron schedule without a human pressing “go.”
She is not a writing tool you operate. She is not a chatbot you prompt per article. Aria is a managed autonomous agent: the production layer that turns research into published content while you focus on strategy and client work.
The distinction matters. Writing tools require you to be in the loop for every piece. Aria requires you to approve the brief and review the finished product. Everything between those two checkpoints runs autonomously.
WEEKLY OPERATIONS
The Production Cycle — How Aria Operates
Aria’s production cycle runs twice daily. Each cycle moves one piece of content from brief to finished draft. The cycle has five stages, and each stage fires on its own cron job so that multiple articles can be in different stages simultaneously.
Morning cycle: Research fires at 7 AM, writing at 8 AM, self-review at 9 AM, art direction runs mid-morning, and the final improve-and-publish stage fires at 10 AM. A brief that enters research at 7 AM can be a finished WordPress draft by 10 AM.
Afternoon cycle: The same five stages repeat starting at 11 AM. This gives the system capacity for two full production runs per day.
Between each stage, the content sits in a queue. If a piece fails a quality gate, it gets flagged and the system moves to the next item. Nothing stalls the line.
| Time | Stage | What Happens | Output |
|---|---|---|---|
| 7:00 AM | Research | Load brief, search knowledge base, run external research, validate citations | Research appendix with verified sources |
| 8:00 AM | Write | Load company context and voice rules, write full draft with SEO metadata | Complete first draft |
| 9:00 AM | Self-Review | Two-pass review: voice/framing check, then grammar/mechanics cleanup | Review report with fix instructions |
| Mid-AM | Art Direction | Generate featured and body images, upload to media library | Images with placement instructions |
| 10:00 AM | Improve + Publish | Apply review fixes, insert images, create WordPress draft, notify for review | Finished WordPress draft with preview link |
| 11:00 AM–2:00 PM | Afternoon cycle repeats all five stages for a second production run | ||
Stage-by-Stage Breakdown
Stage 1: Research. Aria loads the brief, searches the company’s knowledge base and site mirror for existing content on the topic, runs external research via web search, and validates every source URL. The output is a research appendix attached to the brief with verified citations, competitor positioning, and content gaps identified. If the brief doesn’t have enough substance for a quality article, it gets flagged before any writing happens.
Stage 2: Write. With research in hand, Aria loads the full company context: brand voice rules, company information, and target market definitions. She also loads the repertoire tracker, which logs what proof points and frameworks recent articles have used to prevent repetition across the content library. She writes a complete first draft in a single pass, placing internal links, SEO metadata, and image placeholders. Every stat in the draft links to its source.
Stage 3: Self-Review. Aria runs a two-pass review on her own draft. Pass 1 checks voice and framing: are there guru-style structures? Bolded definition lead-ins? Dramatic setups? Teacher-to-student positioning? These are banned patterns, and the review catches and flags each one. Pass 2 checks grammar and mechanics: em-dash overuse, staccato sentence runs, redundant paragraphs, casual markers that aren’t earning their place. The output is a structured review report with specific fix instructions.
Stage 4: Art Direction. Aria generates images for the article using documented art direction rules. Featured images follow a 1200×630px format. Body images are 16:9 landscape. Each image has a specific description tied to the content it accompanies.
Stage 5: Improve and Publish. Aria applies every fix from the self-review, inserts the generated images, creates a WordPress draft with the full content and SEO metadata, enables a preview link, and notifies the content owner for final review. If the piece is low-risk (an optimization to an existing post), it can publish directly. New articles always go to draft status for human review first.
DELIVERABLES
What Aria Produces
Blog Posts
Long-form articles (2,000–4,500 words) written in your documented voice. Each post includes SEO metadata, internal links, image placeholders, and source citations. Published as WordPress drafts for review.
Service & Landing Pages
Full Kadence block pages with structured layouts, FAQ accordions, comparison tables, and CTAs. Designed to match your existing site architecture and visual style.
SEO Metadata
Title tags, meta descriptions, Open Graph tags, and Twitter card metadata set on every piece of published content. Focus keywords placed naturally in H1, H2, and body copy.
Featured Images
AI-generated featured images (1200×630px) and body images (16:9) created to match article content. Uploaded to WordPress media library and set on posts automatically.
Internal Linking
Every article includes contextual links to related pages on your site. Links are placed where they add genuine navigation value, not stuffed for SEO.
Self-Review Reports
A structured review document accompanies every draft, listing specific voice violations found, grammar fixes applied, and brief compliance scores. The report is part of the deliverable, not a hidden internal process.
Page Edits & Optimization
Updates to existing pages and posts: adding new sections, expanding content for keyword coverage, refreshing outdated information. Aria reads the current live page before making surgical changes.
Performance Tracking
Pipeline execution logs track every stage: research sources consulted, review findings, publishing status, and content metrics. Completion reports feed back into brief prioritization for the next cycle.
AGENTIC AI EXAMPLES
Aria in Action
These are concrete tasks Aria performs autonomously. Each fires on schedule, uses specific tools, and produces a defined output.
Brief intake and research. Loads approved briefs from the queue, searches the site mirror and knowledge base for existing coverage, runs external research, validates every citation URL, appends verified findings to the brief. Tools: QMD semantic search, web search, URL validation.
Draft writing with full context. Reads brand voice rules, company information, target market profiles, and the repertoire tracker before writing a single word. Produces a complete draft with headers, internal links, SEO metadata, and image placeholders. Tools: Context file system, WordPress REST API (page retrieval for edits).
Two-pass self-review. Pass 1: voice and framing (25+ banned pattern checks). Pass 2: grammar and mechanics (em-dash audit, staccato smoothing, redundancy elimination). Outputs a structured review report with line-specific fix instructions. Tools: Brand voice ruleset, review template.
Cross-article deduplication. Before publishing, checks the repertoire tracker to confirm no proof points, frameworks, or statistics duplicate what was used in the last three published articles. A reader who encounters five of your articles should find five different reasons to trust you. Tools: Repertoire tracker (JSON), published article index.
Image generation and placement. Creates featured images (1200×630px) and body images (16:9) using AI image generation with documented art direction. Uploads to WordPress media library, sets featured images, and places body images at contextually appropriate locations. Tools: AI image generation, WordPress media API.
WordPress publishing. Builds the full WordPress post or page via REST API, including Kadence block markup for pages. Sets SEO metadata, excerpt, and Open Graph tags. Creates preview links for review. Manages draft/publish status based on risk classification. Tools: WordPress REST API, Kadence block system.
INTEGRATIONS
Systems Aria Connects To
WordPress
REST API for publishing posts and pages, managing drafts, uploading media, and setting metadata.
QMD Semantic Search
Local knowledge base search across company docs, published articles, brand rules, and site mirror content.
AI Image Generation
Creates custom featured images and body illustrations matching article content and documented art direction rules.
Web Research
External web search for competitor analysis, source verification, and industry data gathering during the research stage.
Discord Notifications
Sends completion notifications, preview links, and blocker alerts to your team channel when content is ready for review.
Agent Coordination
File-based mailbox system for receiving briefs from research agents and sending publish notifications to social distribution agents.
Site Mirror
Local markdown copy of every published page, searchable via QMD. Used for internal link discovery and content gap analysis.
QUALITY ARCHITECTURE
How Aria Maintains Quality
Most AI writing sounds like AI writing. That is a solvable problem, but it requires more than prompting for “a professional tone.” Aria uses a documented quality architecture with five checkpoints between brief and published content.
Brand voice calibration. Before writing anything, Aria loads your full voice guide. Banned patterns include guru-style structures, bolded definition lead-ins, dramatic setups, and teacher-to-student framing. Preferred patterns include direct statements, factual transitions, and collaborative framing. Format-specific rules layer on top: blog posts follow different structural conventions than service pages. This is not a generic “write professionally” prompt. It is a documented ruleset with specific examples of what to avoid and what to use instead.
Research quality gate. Before writing begins, Aria checks that the brief’s research section has sufficient substance. Thin research produces thin articles. If the data is not there, the system flags the brief and moves to the next item rather than writing a weak piece.
Two-pass self-review. After the draft is complete, Aria runs two separate review passes. The first pass examines voice and framing, checking against 25+ specific banned patterns. The second pass handles grammar and mechanics: em-dash elimination, sentence rhythm, staccato smoothing, redundancy removal. Each pass produces actionable fix instructions, not vague observations.
Cross-article deduplication. A repertoire tracker logs every framework, statistic, proof point, and company reference used in recent articles. Before publishing, Aria verifies that this piece does not repeat the same evidence as the last three published articles. If your content library grows by 20 articles, each one should present different supporting evidence.
Human review checkpoint. New content always goes to draft status, never straight to publish. A preview link and notification go to the content owner. Edits to existing low-risk content (section additions, metadata updates) can publish directly based on a documented risk classification system.
COMPARISON
Honest Comparison
Different approaches suit different situations. This comparison is straightforward about trade-offs.
| Content Mill | Freelance Writer | Aria (Autonomous Agent) | |
|---|---|---|---|
| Cost per piece | $50–150 | $400–500+ (writing, research, SEO, editing rounds) | $3–5 per article (not per hour) |
| Brand voice match | None. Generic output. | Good after onboarding (2-4 weeks) | Calibrated from day one using documented rules |
| Consistency | High volume, low quality variance (consistently mediocre) | Variable. Depends on their workload and engagement. | Fixed schedule. Same quality process on every piece. |
| Research depth | Surface-level. Recycled talking points. | Depends on the writer. Some excellent, some shallow. | Systematic. Searches knowledge base, site mirror, external sources. Validates citations. |
| SEO integration | Basic keyword stuffing. | Varies. Most writers are not SEO specialists. | Metadata, internal links, keyword placement built into every piece. |
| Self-review | None. | Varies. You are usually the reviewer. | Two-pass review with documented ruleset before you ever see it. |
| Availability | 24/7 ordering, variable delivery. | Business hours. Vacations. Other clients. | Runs on cron schedule. Weekends, holidays, 3 AM — same output. |
| Scaling constraint | Quality floor is too low for B2B authority content. | Cost scales linearly. Hard to maintain 2+ pieces/week. | Adjusting cron schedule and API budget. No hiring. |
When Aria Isn’t the Right Fit
Aria works best for companies that need sustained, consistent content output and have a clear brand voice they can document. She is not the right fit in a few specific situations:
Your content depends on original interviews or proprietary research. Aria can synthesize published sources and your internal knowledge base. She cannot call sources, run surveys, or attend conferences. If your content strategy requires primary research, you need a human writer for those pieces and Aria for the rest.
You prefer to write everything yourself. Some founders want their content to come from their own keyboard. That is a valid preference. Aria is for teams that want the output without the time investment.
You do not have documented brand guidelines. Aria’s quality depends on the rules she loads before writing. During setup, we create these rules from your existing content. But if you have no existing content and no strong opinions about voice, there is not enough signal for calibration. In that case, starting with a human writer who helps establish the voice — and then transitioning to Aria once the patterns are documented — is the better sequence.
TEAM CONTEXT
Part of a Larger System
Aria does not work in isolation. She is one agent in a coordinated system where each agent handles a specific function in the content lifecycle.
Research briefs come from an autonomous SEO research agent that runs keyword analysis, competitive scanning, and content gap identification. Those briefs land in a queue that Aria picks up on her next cron cycle.
After Aria publishes a piece, a notification goes to the social distribution agent, which handles LinkedIn content creation and scheduling. An analytics agent tracks performance across published content and feeds conversion data back to inform future brief prioritization.
This coordination happens through a file-based mailbox system, not APIs or real-time messaging. Each agent reads its own inbox, processes messages, and writes to other agents’ inboxes. Simple, auditable, and resilient to individual agent failures.
ENGAGEMENT
How We Work Together
Setup: Voice Calibration. We document your brand voice, banned patterns, preferred phrasings, and content rules. This is the most important step. The quality of Aria’s output directly reflects the quality of these rules. Calibration typically takes 1-2 weeks and involves reviewing 5-10 existing pieces of your content to extract patterns. Setup runs $100 to $1,200 depending on how well-defined your existing tone of voice is. If you already have a detailed voice guide and strong existing content, the low end covers adapting it for AI-native production. If you are starting from scratch or need help defining your voice, expect the higher range.
Ongoing: Production. Briefs go into the queue. Aria processes them on schedule. You review finished drafts and approve or request changes. Typical turnaround from approved brief to WordPress draft: same day. Your time investment per article averages 5-10 minutes for review, occasionally 15 minutes for pieces that need direction changes. Production cost runs $3-5 per article — better researched, better optimized for search and GEO, with consistent voice calibration on every piece.
Scale: Volume Adjustments. The cron schedule can run more frequently or process multiple items per cycle. Scaling up means adjusting the schedule and API budget, not hiring more people. Scaling down works the same way. Month-to-month, no lock-in contracts.
Aria is deployed through our managed autonomous AI agent service, with typical deployments going live within 30 days.
Frequently Asked Questions
We Have Passionate Customers Who Have Benefited From Our Partnership
Fountain City has been very good at providing the desired results without a lot of hand-holding.
They always have our best interests at heart and come to the table with excellent suggestions based on their experience.
Fountain City offers an excellent personal touch and is very approachable while being very professional still.
Fountain City’s results, honesty and attention to detail stand out. They’re very realistic as well.
I’m always confident they’ll take the lead to ensure that the project runs smoothly and in the manner that I expected.
Results and relationships matter to this team, and it makes for a pleasant and productive experience.
Put Your Content Production on Autopilot
Talk to us about deploying Aria for your content operation. Voice calibration, WordPress integration, and the first test articles are included in setup.
