← All pillars

    Pillar Deep-Dive

    People Ops AI: the guide for People and HR teams

    Where AI fits in the People function, what to build first, and how to keep it running after launch week.

    People Ops AI is the use of AI models, assistants and agents to run People work: answering policy questions, coordinating onboarding, handling hiring admin, drafting documents from approved templates, and turning HR data into decisions. Done well, it moves the People team from a service desk to the system the rest of the company runs on.

    Most People teams start in the wrong place. They buy an assistant, point it at a policy drive nobody has cleaned in years, and wonder why the answers are wrong. The order that works is the one this guide follows: read where the function actually stands, set up a shared workspace, pick one workflow, build it end to end with a person approving the calls that matter, then govern it and measure it.

    Across the People estate there are five domains where AI fits differently: talent acquisition, onboarding and lifecycle, performance and development, operations and compliance, and strategy and insight. The People Ops AI domain map covers each one. Everything below is the full Deepgrain library on People Ops AI, in reading order.

    Skip the reading. The Readiness Assessment scores your People function's AI readiness in about ten minutes.

    Common questions

    What is People Ops AI?

    People Ops AI is the use of AI models, assistants and agents to run People work: answering policy questions, coordinating onboarding, handling hiring admin, drafting documents from approved templates, and turning HR data into decisions. It is a system, not a single tool. The model is the easy part. The data it reads, the actions it can take, the rules on what it never decides alone and the person who maintains it are what make it work.

    What is a People Ops AI assistant?

    An assistant that answers employee and manager questions from your own policies, handbook and HR data, and hands anything sensitive to a person. The useful ones read from a clean policy library, can raise a ticket or update a record rather than only reply, and log every answer so the People team can check them. Pay, performance ratings and disciplinary matters stay with a person.

    Where should a People team start with AI?

    With one workflow, not a platform. Pick a process that is frequent, rule-based and painful, such as onboarding coordination or policy questions, map how it actually runs today, and build the smallest honest version end to end. Once that runs on real data at real volume, the second workflow is cheaper because it reuses the same data access, tools and rules.

    Do we need engineers to build People Ops AI?

    No. You need a champion, a workflow tool, and one clean process. A People person who understands the work and is given time to build will get further than an engineer who does not know how onboarding actually runs. Engineering help matters later, for integrations your HRIS does not expose simply.

    How do you keep People Ops AI safe?

    Write down what AI never decides alone, keep sensitive data inside tools on the right enterprise terms, log what the AI does, and name one owner who reviews it on a set rhythm. Governance done early is what lets the rest run without a nervous manager checking every call.

    Glossary for this pillar

    Terms used across these articles.

    Full glossary →
    AI workspace
    The structured layer of custom instructions, projects, reference documents, and shared prompts that turns a generic AI tool into a function-specific colleague. The first artefact most People teams should build. Read more →
    Champion model
    A staffing pattern for building AI capability inside a function without engineers. Three or four operators given air cover, time, and a build-first remit, supported by a coaching cadence. Read more →
    AI agent
    A model plus a goal plus the ability to take steps. Most companies need three or four agents doing the work that previously clogged three or four roles, not a fleet. Read more →
    AI governance
    The set of rules, escalation paths, audit trails, and human-in-the-loop checkpoints that decide what an AI system is allowed to do. Governance is not the brake. It is the steering. Read more →
    AI workflow
    An end-to-end sequence of work where AI handles one or more steps that previously required judgment. The unit of value most People functions should build against. Read more →
    AI readiness
    The condition of an organisation's data, tools, agents, governance, and cadence such that it can absorb AI as capability. Readiness is not a model selection problem. It is an operating problem. Read more →

    Related across the library

    Keep going.

    Reading is one thing. Doing is another.

    The Grain Audit maps one People Ops process end to end, ranks the highest-return automations, and hands you a 90-day plan you keep whether or not we work together. Two weeks. £2,000, credited in full against a programme. Three slots a month.

    Book a Grain Audit

    Not sure where your function stands yet?Take the Readiness Assessment→