Thready writes, schedules, publishes and measures content across a portfolio of Threads accounts — end-to-end, without humans in the loop. Two operators are running ten accounts today, generating roughly a quarter-million views every month.
Each account is configured once — niche, voice, posting cadence, language. Thready runs the rest as a continuous loop.
A Cerebras-hosted LLM produces drafts using prompts hand-tuned per account — voice, topic clusters, posting style and language are all encoded as templates.
The scheduler picks optimal slots per timezone and per account, posts directly through the Threads API, and confirms delivery before moving on.
Views, likes, replies, follower deltas — collected per post, per account, per day — flow into a single dashboard and an AI assistant you can simply ask.
Pre-wired charts for portfolio views, per-account performance, and content cohorts. Open live dashboard ↗
Each handle is a distinct voice with its own niche, audience and prompt configuration. Performance is tracked independently.
Real-time access to posting stats, top content, and account performance — through a chat interface, not a SQL editor.
The assistant has direct read access to the Thready data warehouse. Ask it anything you would otherwise pull from Superset — it answers in seconds, with the numbers attached.
Single-binary services, embedded analytics database, containerized deploys. We move fast because the substrate is simple.
We're onboarding a small number of additional operators and content partners. The fastest way to reach us is Telegram — usually a reply within the hour.