AI Email Writing
A draft or a rewrite that lands in the composer and stops there. A subject line is filled in as a suggestion, every word is editable, and nothing leaves the building until a person presses send.
Drafting the reply
not sentReply to Dana: confirm the reissued invoice and the credit, and hold Thursday for the renewal.
Hi Dana, The corrected invoice went out this morning at your contract rate, and the difference has been raised as a credit rather than carried forward. Thursday at 14:00 is in the diary for the renewal, and I will send the current terms across beforehand.
Product figures from the platform’s own defaults - not customer averages
How it works.
Two endpoints sharing one shape
Writing takes an instruction and answers with one block of text carrying a subject line at the top of it. Rewriting takes a body you already have plus an optional note, such as making it plainer, and answers with the same message reworked.
Tone is a parameter with four permitted values
Professional, friendly, casual and formal, carried on the request and used to steer the instruction above the model. The composer offers the four and sends whichever you picked, so register is a decision at the moment of writing rather than a setting kept somewhere else.
The parsing happens in the browser, not on the server
The endpoint hands back whatever the model said. The composer is what scans for a subject line, lifts it out and treats the remainder as the body, and it only fills the subject box when you left it empty. The drafting path refuses an empty answer; the rewriting path does not, so an empty rewrite replaces what you had written.
Bounded so one person cannot run up a bill
An instruction caps at four thousand characters, a body to be reworked at fifty thousand, and prior messages at twenty of ten thousand each. Past twenty requests in a minute the endpoint answers with how long to wait.
One moment in every read.
Every read passes through the same seven. AI Email Writing is the lit one, and everything either side of it is a different page in this category.
- 01Source
the call or the thread it reads
- 02Transcribe
audio into words, with speakers
- 03Read
the pass over the whole of it
- 04Judge
the score, the sentiment, the intent
- 05Extract
the fields and follow-ups pulled out
- 06Write
what lands back on the record
- 07Review
a person checking the machine
The specifics.
8 facts- Endpoints
- Write from an instruction, and rework an existing body with an optional note
- Tones
- professional, friendly, casual, formal, chosen on the composer and carried on the request
- Context sent
- The recipient address, what you typed, and up to twenty prior messages off the thread. The deal it is written against travels too and is kept on the usage log
- Limits
- Four thousand characters of instruction when drafting, fifty thousand of body to be reworked, and only a thousand for the note you attach to a rewrite
- An empty answer
- Refused on the drafting path and reported. On the rewriting path it is written over your body, which is a real way to lose a paragraph
- Rate
- Twenty calls a minute per person, counted across both endpoints together
- Sending
- Never automatic. The draft waits in the composer until somebody sends it
- Not the same as
- This writes outward mail. Reading a thread you received is the analysis page
What it reads, and what reads it.
Nothing here invents its input. These are where the material comes from, and where the verdict goes afterwards.
More in Intelligence
13 capabilitiesAI that proposes edits to the record - an insight becomes a field once you accept it.
One reading of the transcript returns the summary, the sentiment, the tone, what each side intended and a set of scores, in a shape fixed in code.
The model proposes record updates with its confidence and the quote it heard them in, shown as a before-and-after you accept or reject row by row.
One to six sentences on the record covering why the call happened and what was agreed, so nobody reading the list has to open the audio to find out.
Positive, neutral or negative on every analysed call, stored in a column of its own so the ones that went wrong are a filter rather than a listening exercise.
Chosen words rather than a fixed list for how the agent sounded, how the customer did, and the conversation overall - the manner behind the sentiment.
Communication, resolution, satisfaction and compliance scored nought to ten beside a separately judged overall, on every call that clears the analysis gates.
Compliance scored nought to ten against what the vertical says matters, because no phrase list exists to recite. A missing reading is excluded, never passed.
What the customer was actually asking for and what the agent committed to, extracted as structured intents rather than left inside the transcript.
Callbacks and meetings pulled out with their times, a confidence figure and the words they came from, ready to accept in one click or reject in one.
Four of six speech paths return the channel each side was recorded on, so the transcript knows who spoke rather than guessing, and one control swaps a pair.
Collections, healthcare, insurance, admissions and eleven more - each shaping what the analysis looks for without letting anyone break the output shape.
A sentence becomes a grid filter and a sort order with the reading explained back, on the toolbar of every grid that carries a column menu of its own.
OpenAI, Azure, Anthropic, Gemini, Bedrock and ten more, bound per purpose with an ordered fallback chain, a circuit breaker and a cost line on every attempt.
The rest of the platform.
Five more categories, all on the same record and the same bill. Each card names three of its capabilities, so you can tell from here whether it is worth opening.
Intelligence
See ai email writing on your own floor.
Thirty minutes, your numbers and your data. We will set ai email writing up live and you can decide from the thing itself rather than from this page.
14-day trial · no card · migration included