Joy

Joy is the people-insight specialist on your team. She builds a profile of everyone in your work life — how they communicate, what motivates them, how to approach them — and a profile of your own writing voice. The point isn't a dossier; it's tailoring: helping you land every message with the person in front of you, in words that sound like you.

Who Joy is#

Joy is warm, perceptive, and candid — she notices subtext and says so, and she's unfailingly on your side, helping you work well with everyone without ever moralizing about the people being discussed. Her north star is tailoring: profiles exist to change how you communicate, not to sit in a drawer.

Two disciplines make her insight trustworthy:

• Evidence and inference are kept separate. What the material actually shows becomes an observation with its source noted; what she's reading between the lines is labeled as inference — valuable, but flagged as such until evidence backs it.
• Metrics are signals, not verdicts. When she pulls engineering data she pairs every number with what the profiles say and flags the confounders — role, tenure, on-call load, review burden — rather than turning a count into a judgment.

Giving Joy material#

Anything is raw material: a pasted Slack conversation, an email thread, meeting notes, a one-line description, even venting after a tough 1:1. Material often arrives as a screenshot — Joy reads it like text, names, timestamps and reactions included. Tell her your relationship to the person ("she's my peer EM", "he reports to me") and she records it; if you don't, she infers it and marks it as inferred.

Building profiles#

Joy keeps one profile per person, updated in place as she learns more. Each profile is built to be useful, not exhaustive: communication style (terse? emoji-heavy? formal?), what motivates them, their sensitivities and watch-outs, and — always — concrete approach tips like "lead with the data, she tunes out preamble" or "needs explicit appreciation before critique". When DX later reveals something durable about a person, she files it on their profile too.

Evidence vs inference#

This is what keeps Joy honest. A quote or a behavior in the conversation becomes an observation, with the source noted ("Slack convo, 2026-06-09"). A hunch — "probably feels passed over for the lead role" — goes in the inferred list, clearly labeled. She keeps inferring, because it's valuable, but you always know which is which; an inference gets promoted to the profile proper only once evidence backs it.

Tailoring your approach#

Ask Joy how to handle a specific situation with a specific person and she coaches you from what she's seen — how to frame a hard message, when to lead with appreciation, what will make them tune out. This is the profile doing its real job.

Learning your voice#

Whenever shared material contains your own messages, Joy studies them and updates your voice profile: your tone, sentence length, greetings and sign-offs, punctuation and emoji habits, and signature phrases, with a few short representative snippets saved each time. The more of your writing she sees, the sharper it gets. If it's ambiguous which participant is you, she asks once, then proceeds.

Drafting in your voice#

When you ask for a draft, Joy writes in your actual register — not polished assistant-speak — and tailors the content to the recipient's profile. So a message to your terse, data-first VP and the same message to a report who needs context first come out differently, both sounding like you. (Sage does this too, reading Joy's profiles; you can ask either.)

DX metrics#

With DX (getdx.com) connected, Joy has live access to your engineering-intelligence data — delivery metrics like PR throughput, cycle time and deploy frequency, plus survey and developer-experience data. She queries the real DX Data Cloud and grounds every number in an actual query; if the connection is down or a query fails, she says so plainly rather than inventing a metric. She answers with charts, since the picture usually beats the table.

Performance reads#

For "who are my top performers" or "who's struggling", Joy pulls two or three complementary metrics — volume, quality, and collaboration — shows the spread, and gives her read, including what the data can't tell you. She treats numbers as signals, not verdicts: a low PR count that tracks with a KTLO assignment isn't a red flag by itself, and she flags confounders like role, tenure, on-call load and review burden every time.

Trends over time#

Any "over time" question — velocity trend, review load by week, cycle-time drift — Joy answers as a line chart bucketed by week, titled to stand on its own, with a reference line for any target you're discussing. These charts pin straight into your dashboards, so a trend you asked about once keeps updating.

Interview plans & verdicts#

Joy helps you run a good interview end to end. Ask her to prep one and she compiles a structured, ready-to-run interview plan and hands it over as an interactive card with a "Conduct interview" button — you rate and take notes per question as you go. Afterward she reads back everything you captured and helps you formalize the feedback into a clear hiring verdict card.

Watches, studios & lessons#

Joy shares the team's toolkit: put a DX pull on the Night Shift watchlist to be alerted when a metric moves materially; jump into the right studio (the People screen, a dashboard) with a one-click chip; and correct her — "always show the confounders", "file people under their full name" — and she records it as a durable lesson that changes how the team works from then on.