Aaptics
Yours Linkedin Digital Twin
We’ve officially hit our first major architecture milestone: The 4-Agent Consensus Engine.
Single-prompt architectures are a liability for high-authority brands. They are unpredictable, prone to hallucinations, and lack the nuance required for a genuine Voice DNA. We decided to stop "prompting" and start Engineering.
What’s under the hood?
Every piece of content generated by Aaptics now undergoes a rigorous multi-agent validation process before it ever touches the scheduler:
The Writer (Inference): Generates the raw structural draft based on the core intent.
The Critic (Quality Control): Scans for "AI-isms," generic patterns, and logical inconsistencies.
The Refiner (Voice Alignment): Cross-references the output with the user’s historical Voice DNA to ensure statistical alignment.
The Orchestrator (Final Delivery): Aggregates the feedback and produces the final, deterministic output.
Why this matters:
Most tools in the market are "probabilistic"—you hit generate and hope for the best. We are moving towards Determinism. By using an agentic workflow, we’ve significantly reduced the "variance" in output quality.
Combined with our Zero-Latency Backend (we’ve eliminated cold starts entirely), Aaptics is no longer just a tool. It’s a mission-critical engine for Founders who value precision over noise.
The architecture of authority is live.
Explore the engine: aaptics.in
Hey Hackers,
The market is currently flooded with AI LinkedIn post generators. But after analyzing the feed, we realized almost all of them are solving the wrong problem.
Generating text is a commodity. The real bottleneck for creators is strategy and data-driven feedback.
A generic AI wrapper cannot tell you if a carousel will outperform a text post today. It does not look at your past analytics to tell you which specific topics actually drive inbound leads.
That is why we built Aaptics.
We are shifting the focus from blind generation to data-driven optimization. Aaptics is a full-stack Creator OS built around what we call the Intelligence Loop.
Here is how our engine works:
1. Data over Prompts: Instead of asking what you want to write, the engine analyzes what is actually working in your specific niche right now.
2. Native Formatting: It crafts the content and formats it specifically for the current LinkedIn algorithm, optimizing hooks, spacing, and pacing.
3. The Feedback Loop: It continuously analyzes the performance of your published posts to dictate your next piece of content. We optimize for revenue and leads, not just vanity metrics.
After a month of rigorous internal testing, we are officially launching publicly today.
We want developers and founders to test this system. We are offering a 10-Day Free Trial to the IH community so you can run the intelligence loop on your own profile.
Check it out here: Aaptics
We will be in the comments all day. Tear apart our UI, test the data loops, and give us your most brutal feedback. We are here to build a real engine, not just another wrapper.
Hey Indie Hackers,
I’ve been increasingly frustrated with the state of LinkedIn. It feels like every second post starts with "In today's fast-paced world" or uses the word "delve." LLMs are great, but default prompting destroys personal branding because it doesn't know your specific syntax, formatting quirks, or vocabulary.
I didn't want to build just another OpenAI wrapper, so I engineered a different approach called Aaptics.
Instead of asking users to write complex prompts to "sound like themselves," I built a "DNA Lab."
Here is how the architecture works:
1. The Extraction Engine (Voice DNA): Users paste 3-4 of their past, human-written posts. The system analyzes the text to extract specific "DNA Clues"—vocabulary frequency, sentence length variation, and formatting styles. It creates a unique profile in the database. When they generate a new post from a raw thought, the AI strictly adheres to this DNA profile.
2. The Tech Stack & Security:
Database: Supabase. I implemented strict Row Level Security (RLS) so users cannot manipulate their API credits or bypass the paywall via the frontend console.
Visuals: I skipped DALL-E 3 because it often ruins text rendering on images. I integrated the Recraft v3 API specifically to generate professional, text-perfect visual concepts alongside the posts.
3. The Scheduling Worker: Handling global timezones was the trickiest part.
The UI takes the user's local time and converts it to a strict UTC ISO string before storing it.
I have a custom Python worker running 24/7. It checks the DB every minute:
if current_utc >= scheduled_utc, it fires the payload directly to the LinkedIn API.
I just launched the V1 and am trying to get my first 10 solid beta testers to break the system and roast the UI.
Questions for the community:
For those who built scheduling tools, is a continuous Python worker loop scalable as user count grows, or should I migrate to a message broker queue (like Redis/Celery) right now?
If you are active on LinkedIn, do you feel authenticity is still a priority, or has the market accepted robotic AI text?
Would love to hear your thoughts, and if anyone wants to test the DNA extraction, let me know!
Explore More - aaptics.in
-
1
How do you evaluate if the DNA extraction truly captures writing style vs the LLM filling gaps with generic patterns?
-
1
That’s the hardest problem to solve! We evaluate the DNA fidelity on three strict vectors:
Negative Constraints: Did it successfully avoid typical LLM crutch words (like "delve" or "tapestry")?
Syntax Matching: Humans write with erratic pacing (a 1-word sentence, then a long paragraph). LLMs write in uniform blocks. We measure if the output rhythm matches the user's historical cadence.
The Zero-Edit Rate: Are beta users hitting "schedule" immediately, or are they manually tweaking the tone?
By injecting the Voice DNA as an isolated, immutable system prompt, we force the LLM to map the raw idea directly into the extracted syntax, leaving no room to hallucinate generic filler.
How do you usually evaluate output quality in your own AI workflows?
-
Comment