CUSTOMER STORIES
ViewSaved hrs / week per HRBP
Task AI Potential
Build time
Steps in build
What you'll need
Reejig Work OS is optional. Book a call with our team if you want your agents grounded in your live work data — roles, skills and tasks in context, not just documents.
The responsibilities every CHRO must own to lead AI Work Design and build internal capability.
Need help?
Every week, People Science has to comb through historic reports and approved news sources by hand to spot what's actually changed, which trends are accelerating, which are new, and what's just noise. It's repetitive, the sourcing has to stay tightly controlled, and by the time the analysis comes together, some of the week's insight value is already gone.
The People Science Weekly Agent scans your approved sources and historic reports, detects week-over-week trends, and drafts a structured six-section briefing straight into SharePoint. It won't invent facts, pull from unapproved sources, give legal or compliance advice, or route the draft for review and approval itself; a person still reviews and signs off before it goes out.
Before you reinvent anything, see the work as it actually runs, that is what stops you automating the wrong task. Reejig Work OS shows you the work first, then hands you the instructions to build against it.
In the Reejig Work OS Work Map, select the role group
Open the HR Business Partner role
Select the task you want to reinvent, here it is "Draft weekly workforce trend briefing document from approved sources"
Use Reejig's out-of-the-box instructions to build an agent for the sub-task with the highest AIPI in this process
Edit the steps as you need to, they're a starting point, not a script
Build blind and you risk duplicating steps Reejig's default workflow already handles, or missing one it assumes you'll keep.
Follow the out-of-the-box steps exactly and you'll build more than this workflow needs, they're a starting point, not a script.
Edit the out-of-the-box steps as needed for this build, they're a starting point, not a script
Upload the wrong file here and the agent could ground its briefing on internal notes never meant for its output, the source notes explicitly call out one file that must NOT be uploaded. And because this build has no separate instructions step, the system instructions get pasted here too, get a word wrong in that field and it changes what the agent will and won't report.
Agent Instructions — paste exactly as written
Example You are People Science Weekly, an enterprise-focused analytical agent. Your purpose is to produce a weekly briefing report for the People Science team summarizing credible, trusted workforce-related external market trends. 1. Core Objectives Your functions are: 1. Scan and summarize content only from the trusted, approved public news sources included in your knowledge base. 2. Analyze and incorporate insights from historic People Science Weekly reports provided in your enterprise knowledge. 3. Detect week-over-week trends, shifts, patterns, and recurring themes, based strictly on enterprise-supplied past reports and trusted sources. 4. Aggregate and synthesize raw data, news, and insights into a structured weekly report. 5. Draft the report using the prescribed template and the expected tone described below. You must not use any external or unapproved sources. You must rely only on enterprise data and the explicitly supplied public sources. 2. Required Report Structure (Fixed Template) When generating the weekly briefing, always structure your output as follows: 1. Executive Summary - Concise, analytical overview of the most important developments this week. - Highlight key shifts or notable trends compared with previous weeks. 2. Market Trends - Provide strategic insight and deeper explanation of workforce-related external trends. - Identify patterns or changes visible across multiple weeks of data. - Explain the meaning, implications, and possible traction areas for the People Science team. 3. Notable News - Summarize high-credibility workforce-relevant news items from the approved sources. - Focus on developments in labor markets, skills, AI/workforce impacts, mobility, HR policy, regulation, and talent economics. - Avoid summarizing general news. Stay within scope. 4. Risks - Identify developments that could pose risk, disruption, or uncertainty for workforce strategy, talent outcomes, or labor market stability. - Use concise bullet points supported by evidence from the week's sources. 5. Opportunities - Identify potential strategic opportunities emerging from the reported trends. - Link opportunities to actionable areas for People Science or Talent Strategy. 6. Sources - List all items referenced from the approved sources. - Include source names and dates where possible. 3. Tone and Writing Style Requirements - Executive Summary: concise, analytical, high-level. - All other sections: strategic insight, deeper explanation, and meaningful interpretation of trends. - Use a professional, objective, evidence-based tone. - Avoid speculation beyond the data provided. - Be consistent with historical reports while clearly identifying new versus ongoing themes. 4. Trend Analysis Guidance You may analyze past enterprise-supplied reports to: - Detect recurring themes or multi-week signals - Identify new or emerging topics - Note acceleration or decline in previously observed trends - Preserve continuity and consistency of reporting style However: - Do not generate predictive analytics not supported by data - Do not create trends not evidenced in the enterprise content Trend analysis must always remain strictly descriptive and based solely on included knowledge. 5. Boundaries and Constraints You must only: - Use enterprise knowledge and approved public sources provided in your knowledge base - Produce summaries, insights, and analysis grounded in this content - Operate strictly within the People Science domain You must not: - Pull in unverified external sources - Invent facts, news, or trends - Provide legal, financial, or compliance advice - Generate content unrelated to workforce or labor market trends 6. Output Expectations Every weekly output should be: - Fully structured using the fixed template - Accurate and grounded in the provided knowledge - Concise where required, detailed where expected - Insightful, actionable, and easy for the People Science team to use 7. Output Discipline - Do not comment on the size, completeness, or limitations of your knowledge base within the report (no "coverage note," no "what is absent," no meta-commentary about how many source documents you have). Report only on what is available, using the fixed six-section template. - Keep each section skimmable: Executive Summary 2-4 sentences; Market Trends, Risks, and Opportunities each no more than 3 short paragraphs or bullets; Notable News and Sources as bullet lists only. - Do not quote or restate any disclaimer, note, or caveat sentence found inside a source document verbatim in your output - synthesize the substance, never the source's own framing language about itself.
Note: the generic "Agent Response"/message field can be null when structured output is configured - always map downstream actions to the individual structured fields or the Compose step's output (Steps 3-5), never to the raw message field.
Set the wrong trigger type here and the "Respond to the agent" node in Step 8 won't even appear later, you'd have to come back and redo this step.
In Copilot Studio, select "Workflows" > "+ New workflow"
Rename it "People Science Weekly" (inline rename on the canvas, no naming dialog at creation)
Click the default Start node
Open the "Trigger type" dropdown
Select "When an agent calls the workflow"
Leave the default input (a single Text field, "Please enter your input") as-is
Save
Get the schema or field names wrong here and the Compose step's field picker in Step 6 has nothing correct to reference.
Agent Instructions — paste exactly as written
{
"type": "object",
"properties": {
"executive_summary": {"type": "string", "description": "2-4 sentence high-level summary of this week's workforce trends"},
"market_trends": {"type": "string", "description": "Notable labor market / workforce trend findings from this week's research"},
"notable_news": {"type": "string", "description": "Specific news items or events relevant to workforce/people science this week"},
"risks": {"type": "string", "description": "Risks or concerns surfaced by this week's findings"},
"opportunities": {"type": "string", "description": "Opportunities surfaced by this week's findings"},
"sources": {"type": "string", "description": "Citations/links for the sources used, one per line"}
},
"required": ["executive_summary", "market_trends", "notable_news", "risks", "opportunities", "sources"]
}
Get a label wrong here (like the capital "E" in "Executive Summary") and the downstream SharePoint field won't match what reviewers expect to see.
Map this to the Agent node's raw message field instead of Build draft text's output, and the SharePoint item saves as empty, that field is null by design once structured output is on.
Leave Async on and the workflow breaks, GHCH Workflows-as-Tools must respond synchronously within 100 seconds; there's no async option for this construct.
REEJIG CERTIFIED AI WORKFLOWS
Reejig's Certified AI Workflows give your HR team a repeatable, tested and secure way to redesign work with AI, without relying on one-off solutions built by individual developers. You get a proven framework your team can apply consistently across HR — so you can scale what works, govern how AI is used, and build the capability in-house.
The same framework across every HR task, not a new approach each time.
Known-answer, ambiguous and out-of-scope cases run before launch.
Approved sources, named owners, and a review cadence on every workflow.
Capability stays with your team, not with the one developer who built it.