Classifies hundreds of emails with AI, surfaces the few that genuinely need a reply, suggests a draft for simple/social emails (never sends), and produces a daily digest. No subscription — pay per email processed.
Input
| Input | Type | What it does |
|---|---|---|
emails |
array | The emails to triage |
kullanici_isimleri |
array | Emails that mention your name/surname are flagged as 'addressed to you' and surfaced |
kullanici_eposta |
string | By looking at the To/Cc headers: if you are only CC'd (not the direct addressee) the email is treated as 'informational' and deprioritized |
language |
string | The language of the classification text and draft replies |
model |
string | Classification model |
process |
boolean | When off, only a FREE preview (email count + how many can be filtered without the LLM) |
Running it
Start it from the Apify Console, call it over the REST API, or expose it to an agent as an MCP tool — the dataset is the same either way.
What it costs
Pay per event, not per hour: starting a run is free, and you are charged for the results you actually get.
Top comments (1)
The approach of classifying emails based on their relevance and priority is a crucial innovation, especially with the overwhelming volume we face daily. Utilizing user-specific flags like
kullanici_isimlerito prioritize messages is a smart way to enhance user experience, as it directly aligns with how we typically manage our inboxes. One potential improvement could be incorporating user feedback for the AI to learn which classifications work best over time, allowing for more personalized triage results. If you're looking for assistance in refining the classification model or expanding its capabilities, I’d be interested in discussing a paid collaboration to explore those enhancements further.