Get help preparing a conversation, drafting a goal or finding the next action. October’s AI assistance lives inside people workflows with role-based access and human review.
Help me prepare for my next 1:1 with Naledi.
Start with the person, then the goal. Naledi’s product work gives you a useful opening for a conversation about leading the next discovery initiative.
Discuss a small discovery initiative Naledi could lead this month.
Illustrative responses · Review suggestions before taking action
Who it’s for
October’s AI is most useful where people already make decisions. These are the moments we built it for.
One person supporting four hundred employees across several teams.
“I find out someone was unhappy when they hand in their resignation.”
Flight-risk predictions combine tenure, pay, performance, manager changes, 1:1s, recognition and leave. Higher scores create follow-up actions for the manager or HR.
Eight direct reports and back-to-back meetings.
“I walk into 1:1s with nothing prepared and leave without notes.”
1:1 prep briefs pull together goals, previous actions and recent feedback. Recaps capture what was agreed, and nudges flag who hasn’t had a 1:1 or recognition lately.
Calibration across managers who rate very differently.
“A rating depends more on who your manager is than on what you did.”
A fairness check reviews each calibration session for manager consistency, outliers and demographic patterns, and suggests rating adjustments with the evidence.
Planning next year’s headcount and hiring budget.
“How many people will we lose next year, and where?”
Attrition forecasts start from your own history, adjust for current flight risk and add retirement projections, so the plan reflects the organisation as it is today.
Why people data goes unused
TodayRetention risk is a gut feeling, or an exit interview after the fact.
With OctoberFlight-risk scores with contributing factors, and suggested next steps such as a stay interview or a pay review.
TodayManagers prepare for 1:1s by scrolling back through old notes, if at all.
With OctoberA prep brief and talking points arrive before the meeting, with open actions carried forward.
TodayCalibration bias is spotted by whoever happens to be in the room.
With OctoberEvery session can be checked for manager consistency, outliers and demographic patterns.
TodayPromotion cases are judged on how well they’re written.
With OctoberAI flags vague or unsupported claims and missing evidence, so cases are judged on the work.
AI in the work
Four places where October’s AI turns people data into something a person can act on. Each one shows its reasoning and leaves the decision with you.
Retention signals
Flight-risk predictions score each employee from 0 to 100 using tenure, compensation, performance, manager changes, 1:1s, recognition and leave. Each score lists the factors behind it, sorted by impact, with a confidence level.
Forecasting
Attrition forecasts project monthly departures from your own history, then adjust for the number of people currently at high flight risk. Tenure bands and retirement projections show where the losses are likely to come from.
Most departures in the last year came from people in their first two years. Retirement projections add 3 expected exits next year.
Fair decisions
Before ratings are final, a fairness check reviews the session. It looks for managers who rate consistently high or low, ratings that don’t match the evidence and demographic patterns, and suggests adjustments for the panel to discuss.
Ratings are broadly consistent. One manager rates noticeably higher than the session, and one rating looks out of line with goal results.
Rating 2 → 3. Goal evidence and peer feedback point to solid performance. The written review focuses on one missed deadline.
Assistance in workflows
In promotions, AI can draft evidence from goals and feedback, flag criteria with little support and check language for vague or exaggerated claims. It prepares a neutral summary for the panel. It never scores readiness or ranks candidates.
Kwame went above and beyond on the billing migration and single-handedly doubled team output. He now runs the weekly incident review and mentors two junior engineers.
Influences beyond the team. Little supporting evidence in goals, feedback or recognition.
Suggestions only. Each use is recorded on the case.
From signal to conversation
An illustrative engineering team at a 480-person company. The aim is a better conversation, not a prediction that comes true.
A flight-risk prediction run scores Thandi at 68, high risk. The top factors are below-market pay, no raise in 22 months and a recent manager change.
Creates a manager alert and a stay-interview prompt for her manager, due in 14 days. The HR business partner sees both on the interventions list.
Opens the prep brief for Thursday’s 1:1. It suggests talking points and flags an action he still owes her from last time.
Holds the conversation. The recap captures her interest in a certification and a question about pay.
Reviews her pay against the band ahead of the compensation cycle, then marks the intervention complete.
Less guessing. Earlier conversations.
How it works
Records, reviews, goals, 1:1s, recognition and leave live in October People, so AI works from the same picture HR sees.
Turn on the AI features that suit your organisation. Access follows each person’s role.
Read the reasoning behind each signal or suggestion, then decide what happens next. Nothing changes without a person.
Human oversight
Flight-risk scores, calibration checks and suggestions show the factors and evidence behind them.
AI never makes a rating, promotion or pay decision. Suggestions wait for a person to apply or ignore them.
AI use on promotion cases is recorded for the audit trail, and AI features can be turned off for your organisation.
Models are listed in October’s published register, and customer data is never used to train them.
Where it connects
Calibration fairness checks and review summaries sit inside review cycles.
ExplorePrep briefs, talking points and recaps for every 1:1.
ExploreFlight risk, high potentials and succession readiness in one view.
ExploreAttrition forecasts feed headcount and hiring plans.
ExploreExplore the workflows, tools and practical details that support ai assistance throughout the working day.
Timely, contextual nudges: "Schedule a check-in with Sarah." "James hasn't received feedback in 30 days." "Team morale is trending down." Recognition gaps, overdue 1:1s and engagement dips reach managers in time, and an AI assistant answers any question about the platform.
An AI career coach helps every employee explore trajectories, identify skill gaps and build development plans tailored to their ambitions.
An AI high-potential score built from performance, potential, learning, goals and growth. Succession readiness scoring, what-if vacancy simulations and machine-learning flight-risk predictions with explainability.
Cara sources and scores candidates and starts the outreach. AI screens applications against your criteria, runs voice interviews and reports how well its recommendations matched hiring outcomes.
AI transcribes 1:1s and key meetings, extracts action items and analyses sentiment over time. Generate goals, job descriptions, policies and announcements from a sentence. Personality insights add work-style context to assessments.
Named purpose, access controls and human oversight for every model.
Every model, purpose and control listed in the October model register.
Explore the detail behind ai assistance.
AI works inside governed workflows with access controls, named purpose and human oversight. Customer data is never used to train models. October publishes its model register, governance policy and evaluation results.
Yes. Each AI capability can be enabled or disabled per organisation, and outputs are always reviewable by a person before they become a decision.
We’ll walk through the real product with your team and show how it connects to the rest of October People.