Eagentix Hands-On AI Training

HR & People AI Training Singapore

Hands-on AI training in Singapore for HR and people teams on Copilot, ChatGPT, Gemini, Claude or Amazon Q, using your own policies and templates.

AI training for hr & people teams

AI training for HR and people teams

HR teams are being given AI assistants without being shown how to use them on people work, where the cost of a confident wrong answer is high. This is a hands-on programme for HR business partners, talent acquisition and people operations teams who need to apply Microsoft Copilot, Claude Cowork, Gemini Enterprise, Amazon Q Enterprise or ChatGPT Enterprise to recruitment preparation, onboarding, policy questions and people reporting — with the checks that keep decisions about people with people. Sessions are delivered in Singapore, on site or remote, using your tool licences and anonymised versions of your own documents.

Who this programme is for

HR, talent and people operations teams.

  • HR business partners who answer the same policy questions every week
  • Talent acquisition teams preparing role profiles, interview questions and assessment rubrics
  • People operations teams running onboarding and employee communications
  • HR leaders deciding where AI may support people decisions and where it must not

Who it is not for

  • Teams looking for a vendor demo rather than hands-on practice
  • Anyone expecting AI to screen out, rank or decide on candidates or employees without human judgement
  • HRIS implementation or system integration projects — this is applied assistant training

Assistants covered in the programme

The programme is delivered on the assistant your organisation already licenses. The techniques transfer between them.

Microsoft Copilot

Word, Outlook, Teams and Excel inside Microsoft 365. Used for role profiles, interview guides, onboarding packs and policy answers drafted from documents you already hold.

ChatGPT Enterprise

Analysis over uploaded files, custom GPTs for repeat HR tasks such as policy Q&A, and prompt patterns for people reporting.

Gemini Enterprise

Google Workspace HR work: Docs for role profiles and onboarding plans, Sheets for workforce reporting, and grounded answers over handbooks in Drive.

Claude Cowork

Long policy sets, employment agreements and handbooks, with answers that name the governing clause so they can be checked.

Amazon Q Enterprise

Grounded answers over connected HR sources for organisations standardised on AWS.

Curriculum and hands-on modules

Module 1: Recruitment preparation

Hands-on exercise. Draft role profiles, structured interview questions and competency rubrics for human-led assessment.

What you leave with. A reusable hiring preparation pack.

Module 2: Onboarding and policy

Hands-on exercise. Build an onboarding plan and draft policy answers grounded in approved documents.

What you leave with. An onboarding checklist and source-backed policy response template.

Module 3: People reporting

Hands-on exercise. Summarise anonymised feedback and prepare workforce reporting with explicit checks for omissions and bias.

What you leave with. A reviewed people report and quality checklist.

The same hr & people day on each assistant

The modules stay the same; the exercise changes with the tool your team is licensed for.

Microsoft Copilot: Structured interview guide in Word

What goes in
An existing job description and your competency framework.
What is asked
Draft two behavioural questions per competency, with a note on what strong and weak evidence sounds like for each.
What comes out
An interview guide panel members can use as written.
What the reviewer checks
Every question maps to a named competency, and nothing touches age, family status, race or religion, in line with Singapore's fair employment guidelines.

ChatGPT Enterprise: A policy-answering GPT on your handbook

What goes in
Your approved employee handbook and leave policy.
What is asked
Build a GPT that answers staff policy questions, quotes the section it relied on and declines when the handbook is silent.
What comes out
A shareable GPT plus a ten-question test set.
What the reviewer checks
Three of the ten test questions are deliberately outside the handbook; the GPT passes only if it declines all three.

Claude Cowork: Employment agreement comparison

What goes in
Your current employment agreement template and three older versions still in force.
What is asked
Compare notice periods, restraint clauses, leave entitlements and benefits across all four, naming the governing clause each time.
What comes out
A comparison table showing which staff are on which terms.
What the reviewer checks
HR confirms each difference against the signed originals before anyone is told their terms differ.

Gemini Enterprise: Engagement survey themes in Sheets

What goes in
Anonymised free-text survey responses in a Sheet.
What is asked
Group responses into themes, count each, and list the responses that fit no theme.
What comes out
A theme table and a short Docs summary for the leadership readout.
What the reviewer checks
The unthemed list is read in full, and small teams are aggregated so no individual can be identified.

Amazon Q Enterprise: Policy answers from connected handbook content

What goes in
The connected HR handbook source.
What is asked
Answer common staff questions, citing the document used, inside your existing permissions.
What comes out
A tested question set with sources.
What the reviewer checks
The team checks which documents the connection can and cannot see before relying on an answer.

Singapore rules the exercises respect

HR material is the most sensitive data most organisations hold, and Singapore law sets clear limits on how it is used.

  • Employee personal data is covered by the PDPA, including purpose limitation: data collected for payroll is not automatically available for an AI exercise.
  • Hiring content is reviewed against the Tripartite Guidelines on Fair Employment Practices and the Workplace Fairness Act passed in 2025.
  • Survey and feedback analysis is aggregated so that small teams cannot be identified.
  • Decisions about individuals stay with people. The programme covers preparation and review, not automated screening.

How the sessions run

Delivery

On site in Singapore or remote, delivered live. Cohorts are kept small enough that every participant works on their own files rather than watching a screen share.

Duration and scheduling

Typically one full day, or two half-days scheduled around hiring and review cycles. Module order is adjusted to your calendar and tool stack.

Session structure

Each module follows the same pattern: a short walkthrough of the technique, a hands-on exercise on your own data, then a review of what the assistant got wrong and how to catch it.

After the session

Participants leave with the prompt patterns, templates and review checklists produced during the session, so the work continues after the training day.

Prerequisites

  • A licensed seat on at least one of the covered assistants (we can run the programme on whichever your organisation has)
  • Working knowledge of your own HR processes, handbook and policy library
  • Anonymised or sample working files — a role description, a policy extract, an onboarding checklist
  • No coding background required

Data handling and human review

  • Exercises run inside your own tenant and your own licences, so your data does not leave your environment for the training.
  • Files shared with the trainer in advance are anonymised extracts; employee records, candidate data and case files stay inside your own tenant and licences.
  • Every module includes verification: checking output against the source, confirming what the assistant left out, and marking what still needs a human read.
  • Decisions about candidates and employees stay with your people. The programme teaches AI-assisted preparation and review, never automated hiring or employment decisions.

Who delivers the training

Abhijit Ghosh

The programme is delivered by Abhijit Ghosh, who has spent 25 years in data and AI, including roles at Accenture, IBM, Deloitte, Cloudera, Tibco and Wipro. Sessions are run by the practitioner, not handed to a junior facilitator.

Frequently asked questions

Which AI assistant should our HR team use?

Start with the one your organisation already licenses. Microsoft Copilot suits teams working in Outlook, Word, Excel and Teams; Gemini Enterprise suits Google Workspace; ChatGPT Enterprise and Claude Cowork are stronger for long, document-heavy work; Amazon Q suits AWS-standardised estates. The programme runs on whichever you have, and the techniques transfer between them.

Do participants need a technical or coding background?

No. The programme is built for practitioners, not engineers. If you can describe how the work is done today and you know your own documents and systems, you have the prerequisites. Nothing in the sessions requires scripting.

Will we work on our own material?

Yes. Private sessions run on your own handbook, role descriptions and templates inside your own tenant. Where a file has to be shared with the trainer in advance, an anonymised extract is used, and personal data about candidates or employees is never needed for the exercises.

How long is the programme and how is it scheduled?

Typically one full day, or two half-days. HR teams usually schedule it away from peak hiring or performance-review periods. Module order is adjusted to your tool stack and calendar.

Is the training delivered in Singapore?

Yes. Sessions are delivered on site in Singapore, or live and remote for teams spread across the region.

What do participants take away?

A reusable hiring preparation pack, an onboarding checklist with a source-backed policy response template, and a reviewed people report with a quality checklist — all produced during the exercises using your own material.

How is this different from a vendor product demo?

A demo shows features on prepared data. This programme applies the assistant to your own work, including the cases where it produces confident, wrong output — because recognising and catching those is the skill that matters once the team uses it unsupervised.

Can an AI assistant be used to screen candidates?

It can prepare the material a recruiter reviews — structured role criteria, interview questions, a summary of what a CV does and does not show against those criteria. The assessment and the decision stay with a person. Every recruitment workflow in the programme makes that human decision point explicit, which is also what employment regulators expect.

How does the programme handle PDPA for employee data?

Exercises use anonymised or synthetic records unless your data protection officer has cleared real data for the purpose. Purpose limitation is covered explicitly, because data collected for payroll or performance is not automatically available for an AI exercise.

Can the assistant screen candidates?

The programme does not teach automated screening. It covers drafting role profiles and interview guides and reviewing them against fair employment guidelines. Decisions about candidates stay with people.

Choose the assistant your team already licenses

HR & People training by assistant