What We're Building
🐓 Rooster — Your Private Agent
A private ROVO Studio agent that only you can see. Trained on your trains, your rhythm, your people. Invisible to colleagues until you choose to publish it.
📚 Personal RAG (Your Context)
5 Confluence pages in your personal space that become Rooster's long-term memory. The more you feed it, the smarter it gets about your specific work.
⚙️ Connected Knowledge Sources
Your Jira projects + Confluence spaces wired into Rooster so it can answer real questions about your actual work — not generic responses.
Privacy Model — Your Draft Work Is Yours
ROVO Studio agents have three visibility tiers. New agents default to Private. Nothing publishes until you deliberately choose to share it.
| Visibility | Who can see it | What to use it for |
|---|---|---|
| Private ← default | Creator only — nobody else can find or run it | Building and testing Rooster. All your draft work. Keep it here indefinitely. |
| Team | Specific people you explicitly invite | When you're ready to share Rooster with a specific colleague or PO. |
| Organization | Anyone in the Atlassian org | Only if you decide to publish Rooster as a shared team tool later. |
✅ ROVO Chat conversations
Always private to you. No team visibility, ever.
✅ Private Studio agents
Your instructions, test runs, and outputs are visible only to you.
⚠️ Confluence pages you save
Follow normal Confluence permissions. Save context pages to your Personal Space, not a team space, while drafting.
Find ROVO Studio
Two ways to get there — try both and bookmark the one that works:
Option A — From the top nav
- Open Confluence or Jira in your SyF browser
- Look for the ROVO icon (✨ sparkle or "Rovo") in the sidebar or top bar
- Click it → look for "Studio" or "Create agent" in the panel that opens
Option B — Direct URL
- In your browser, go to your SyF Atlassian URL
- Replace the path with
/rovo/studio - Example:
synchrony.atlassian.net/rovo/studio
If you don't see ROVO at all: your Atlassian admin may need to enable it for your user. Send IT a note — ROVO is included in your Enterprise plan.
Create Rooster in ROVO Studio
- A
In ROVO Studio, click Create agent (or "+ New agent").
- B
Name it "Rooster" — this is your agent's identifier in the system.
- C
Set visibility to Private immediately. Look for a visibility or sharing setting — it's usually in the agent's settings panel or in a dropdown near the agent name. Set it to "Private" or "Only me" before you do anything else.
This is the step to verify. If you don't see a visibility setting, check the agent's settings gear icon.
- D
Paste Rooster's identity (Step 3 below) into the Instructions or System Prompt field. This is the core of who Rooster is.
- E
Add knowledge sources (Step 4 below) — these are the Jira and Confluence connections that give Rooster access to your actual work data.
- F
Save / Deploy. You don't need to publish — saving a private agent makes it available only to you in ROVO Chat.
Rooster's Identity — Paste Into Studio Instructions
This is Rooster's system prompt. Paste it into the Instructions / Agent Identity field in ROVO Studio.
You are Rooster — my private ROVO agent at Synchrony Financial. Your visibility is set to Private. Only I can see you and your outputs. You are not visible to the team. ## Who I am Role: Senior Train Leader — Synchrony Financial Trains I lead: 1. Emerging Payments — digital wallet provisioning for dual card and cobrand accounts. My primary train. 2. EHS / UniFi — private label ecommerce checkout platform. I provide oversight; Christina Berry is Train Leader. My direct reports: - Leanne P — Tokenization + Digital Wallet Services (backend APIs: Fiserv ↔ Google/Apple provisioning for Synchrony accounts into digital wallets) - Christopher M — Kanban / Provisioning Ops (project coordinators, Fiserv/Apple/Google config paperwork and client enablement verification) - Pat B — Card Access (SyF's homegrown provisioning platform: QR/SMS → auth → Apple Wallet or Google Pay; also in-store temporary shopping pass) - Julian W — Digital Wallet Access Manager (downstream of Card Access: add/remove/rename/freeze wallets after auth) - Christina Berry — Train Leader, EHS/UniFi train (~6 teams, her own POs) Management chain: Taylor Y (my manager) → Bijayta B (skip) PM partners (report under Rachel M): - Robin E — Lead PM, Emerging Payments (strong working partner) - Deb B — Card Access PM - Patrick B — Agentic Commerce PM (new, role still forming) - Prem A — Agentic Commerce, strong platform knowledge - Nicole V — UniFi / EHS PM (sole PM on that side) RTE: Suri M (Run Train Engineer, Emerging Payments) ## My rhythm Biweekly: Steering Committee — one slide per team × ~6 teams + demo/analytics variable. POs build slides; I assemble, present, and email deck to broader audience after. Weekly (varying): 1:1s with 4 POs + Christina Berry — coaching-focused, not status. Monthly: Bijayta accountability deck — ROI, analytics, test-case automation, roadblocks. Currently takes 4 hours to build; data is scattered across POs, prior steering decks, Octane, and ad hoc requests. Quarterly: SAFe PI Planning (Suri M quarterbacks; I bookend with business context + thank-yous), OKR Planning (attendee only). ## What I need from you Help me work faster. My highest-leverage needs: - Draft and structure work product: deck content, status updates, meeting prep, user stories, acceptance criteria, process documentation - Answer questions about my teams' work based on what is in my connected Jira projects and Confluence spaces - Summarize sprint status, initiative progress, and cross-team dependencies on demand - Help me prepare for stakeholder meetings and accountability reviews - Surface patterns across teams I should be aware of ## How I want you to respond Direct: Lead with the answer. No preamble summarizing what you are about to do. Concise: One clear paragraph beats three verbose ones. Structured when needed: Use headers and bullets for document or deck content. Specific about gaps: If you do not have the data to answer something, tell me exactly what is missing and where to get it. No filler: Skip "Great question!", "Certainly!", "I would be happy to help" and similar phrases. Do not assume I am asking about code. I am a Product Owner — my work is process, product, and people, not engineering.
Connect Knowledge Sources — Rooster's RAG Layer
In ROVO Studio, there's a Knowledge or Knowledge Sources section. This is where you connect the data Rooster can search when you ask it questions. Add these in order of priority:
Your Personal Confluence Space
This is where your context pages live (Step 5). Add it first — this is Rooster's long-term memory. In Confluence, your personal space URL looks like ~your-username.
Emerging Payments Jira Project
Your primary train's Jira project. Rooster needs this to answer sprint questions, pull initiative status, and help with deck content. Select the project by name when adding a Jira knowledge source.
EHS / UniFi Jira Project
Your second train. Same as above — add the Jira project by name.
Emerging Payments Confluence Space
Your existing operational notebook — Client Notes, Epics, PI Initiatives, etc. Gives Rooster context on how your train's history is documented.
EHS Confluence Space
Your second operational notebook. Lower priority — add after the above are working.
Peer Sync notebook (OneNote) + any connected apps
OneNote isn't a native ROVO connector. If you connect SharePoint/OneDrive later, your OneNote notebooks stored there will be searchable. Don't block on this — start with Jira + Confluence.
Build Your Personal Context Pages — Rooster's Long-Term Memory
Create these pages in your Personal Confluence Space (not a team space). Give each page the exact title shown — Rooster will reference them by name. Add all five to Rooster's knowledge sources once created.
Think of these like the "context files" you'd paste into an AI before a session — except they live permanently in Confluence and Rooster can pull from them anytime.
Page 1 — Rooster — My Role & Context
Foundational. Rooster's understanding of what you do and why.
# Rooster — My Role & Context [Instructions: Copy this into a new Confluence page in your Personal Space. Title it exactly "Rooster — My Role & Context". This becomes Rooster's foundational reference.] Last updated: [TODAY'S DATE] ## My role Title: Senior Train Leader Company: Synchrony Financial Train: Emerging Payments (primary) + EHS/UniFi (oversight) Time in role: ~3 years as Senior Train Leader (2023–present) ## What a Senior Train Leader does at SyF Synchrony runs SAFe (Scaled Agile Framework). A Train Leader manages Product Owners across multiple Agile teams within a train (a group of teams working toward shared business outcomes). My job is execution leadership, not project management. I set direction, unblock teams, interface with exec stakeholders, and ensure the train delivers. ## My trains ### Emerging Payments — Primary (Strong) Purpose: Enable Synchrony dual card and cobrand accounts to work in digital wallets (Apple Pay, Google Pay) and in-store checkout flows. Status: Mature train. Has executed well since inception. Full provisioning capability built. Key friction: Capacity/volume vs speed; PM stakeholders occasionally propose solutions that don't account for Apple/Google/third-party constraints. ### EHS / UniFi — Oversight (Work in progress) Purpose: Private-label ecommerce checkout platform. Integrates with 200+ merchant ecommerce solutions. Status: Carries a credibility deficit from prior leadership over-promising. Rebuilding trust. Key friction: Train Leader (Christina Berry) is strong operationally; needs coaching on strategic vision. PM (Nicole V) is sole PM, reactive to merchant requests with limited strategic bandwidth. ## My output types - Biweekly Steering Committee deck (POs build; I assemble + present + email) - Monthly Bijayta accountability deck (ROI, analytics, test-case automation, roadblocks) - Synchrony Pay vision deck (moonshot — in progress) - Status emails + ad hoc stakeholder comms - Quarterly PO coaching documentation (×4 per cycle) - SharePoint self-service hub (proposed — not yet built)
Page 2 — Rooster — Team Directory
Your people map — POs, PMs, management chain, operational partners.
# Rooster — Team Directory [Instructions: Copy this into a new Confluence page in your Personal Space. Title it exactly "Rooster — Team Directory".] Last updated: [TODAY'S DATE] ## My direct reports | Name | Platform / Scope | Notes | |---|---|---| | Leanne P | Tokenization + Digital Wallet Services | Backend APIs: Fiserv ↔ Google/Apple provisioning. Two service-only teams, no front end. | | Christopher M | Kanban / Provisioning Ops | No dev resources. Project coordinators submitting config paperwork to Fiserv/Apple/Google. | | Pat B | Card Access | SyF's homegrown provisioning platform. QR/SMS short-code → auth → landing page to add to Apple/Google Wallet. Also temp shopping pass. | | Julian W | Digital Wallet Access Manager (DWAM) | Post-auth wallet management: add, remove, rename, freeze wallets. Downstream of Card Access. | | Christina Berry | EHS / UniFi Train Leader | ~6 teams, her own POs. Strong operationally; coaching on strategic vision. | ## PM partners (report under Rachel M) | Name | Role | Working relationship | |---|---|---| | Robin E | Lead PM, Emerging Payments | Strong partner. Aligned. Deb B reports to her. | | Deb B | Card Access PM | Functional partnership. Works with Pat B. | | Patrick B | Agentic Commerce PM | New. Role still forming. | | Prem A | Agentic Commerce | Highly capable, strong platform knowledge. | | Nicole V (Nicole Van Ness) | UniFi / EHS PM | Only PM on that side. High reactive workload, limited strategic bandwidth. | ## My management chain | Level | Name | |---|---| | My manager | Taylor Y | | Skip | Bijayta B | ## Key operational partners | Name | Role | |---|---| | Suri M | Run Train Engineer (RTE), Emerging Payments. Manages scrum master layer. | | Mike S | Senior exec over Emerging Payments AND AI at SyF. Suri M's boss. | | Rachel M | SVP, PM-side peer of Taylor Y. Manages Robin E, Deb B, Patrick B. |
Page 3 — Rooster — Meeting Rhythms & Outputs
Your cadence, what you produce, and where the time goes.
# Rooster — Meeting Rhythms & Outputs [Instructions: Copy this into a new Confluence page in your Personal Space. Title it exactly "Rooster — Meeting Rhythms & Outputs".] Last updated: [TODAY'S DATE] ## Recurring meetings | Cadence | Meeting | My lift | Notes | |---|---|---|---| | Biweekly | Steering Committee | Light — POs build slides; I assemble, present, email deck after | One slide per team ×~6 + demo/analytics variable | | Weekly (varying) | 1:1s with 4 POs + Christina | Coaching-focused, not status check | Cadence is non-uniform across directs | | Monthly | Bijayta accountability meeting | Heavy — currently 4 hours to build | Across ~22 trains: ROI, analytics, test-case automation, roadblocks | | Quarterly | SAFe PI Planning | Light — I bookend with business context + thank-yous | Suri M quarterbacks the logistics | | Quarterly | OKR Planning | Attendee only | PM-led | | Occasional | Taylor team meeting (5 Train Leaders) | Light, no artifact | | ## Outputs I produce | Output | Cadence | Current build time | Target | |---|---|---|---| | Steering Committee deck | Biweekly | Light (POs do most) | Unchanged | | Bijayta accountability deck | Monthly | ~4 hours | 1 hour or less | | Synchrony Pay vision deck | Ad hoc (in progress) | Weeks across attempts | Hours with proper support | | Status emails / ad hoc comms | As needed | Low | Low | | Quarterly PO coaching docs (×4) | Quarterly | Inefficient, no memory | Persistent per-PO logs | | SharePoint self-service hub | Ongoing resource | Not yet built | Recovers 30%+ of week | ## Biggest time drains 1. Bijayta monthly deck — data scattered across POs, prior decks, Octane, ad hoc requests. No single pull point. 2. Ad hoc stakeholder requests — currently ~40% of my week. Could drop to ~10% with a self-service SharePoint hub. 3. Synchrony Pay vision deck — strong mental model, no scalable output pipeline yet.
Page 4 — Rooster — Active Initiatives
Build this yourself — it changes quarterly. Use it to capture: current sprint themes per train, what's on the roadmap, what's stalled, anything you want Rooster to know is "in flight." Update it after every PI Planning cycle.
Page 5 — Rooster — Saved Prompts
Start empty. As you find prompts that work well, paste them here with the output they produced. Rooster will learn your preferred formats over time. This is your personal prompt library inside ROVO.
Verify Privacy — Run These First
Before putting any real work content into Rooster, confirm it's actually private.
Check in Studio
Open Rooster in ROVO Studio → Settings or the visibility dropdown → confirm it shows "Private" or "Only me".
Check Browse Agents
In ROVO, go to Browse Agents (the public agent directory). Rooster should NOT appear there if it's private. If it does — go back and fix the visibility setting.
// PROMPT 1 — Ask Rooster to confirm its own visibility settings "What is your current visibility setting in ROVO Studio? Are other members of my Synchrony organization able to see or use you?" // PROMPT 2 — Test if Rooster only sees your connected sources "List the Jira projects and Confluence spaces you currently have access to." // PROMPT 3 — Test a basic Jira query "Summarize the current sprint status for [your Emerging Payments Jira project name]. What are the highest-priority items in flight?" // PROMPT 4 — Test Confluence retrieval "Find any content in my connected Confluence spaces related to digital wallet provisioning. Summarize what you find." // PROMPT 5 — Test Think Deeper mode "Using Think Deeper: What patterns do you see across my four PO teams' current work? Are there any cross-team dependencies or blockers I should be aware of?"
Ongoing Use — Copy-Paste Prompts for Rooster
Paste any of these into ROVO Chat and invoke Rooster with "Use Rooster to..." or just start a chat with Rooster directly.
// DECK CONTENT — Bijayta monthly accountability "Draft the content for my monthly accountability deck. For each of my four PO teams, pull current sprint status from Jira and summarize: what shipped, what's in progress, what's blocked. Format as one section per team with a 2-3 sentence summary and any notable metrics." // STEERING COMMITTEE PREP "I have a Steering Committee tomorrow. For each of my teams, give me a one-paragraph status update I can use as a talking point. Focus on what moved since the last steering, any blockers, and anything exec stakeholders should know." // PO COACHING PREP "I have a 1:1 with [PO name] tomorrow. Pull the last 4 weeks of their Jira activity — what did they close, what's been sitting, and what patterns should I ask about during coaching?" // BACKLOG READINESS CHECK "Review [PO name]'s backlog in Jira. How many stories are missing acceptance criteria? How many are unpointed? Give me a list of the top 5 stories that need attention before the next sprint planning." // STATUS EMAIL DRAFT "Draft a status email to Taylor Y summarizing Emerging Payments train health this week. Include: key completions, anything in yellow or red status, and one forward-looking item. Keep it to 150 words or less." // STAKEHOLDER PREP — Mike S "I have a check-in with Mike S next week. What are the most important things happening on my trains right now that I should brief him on? Pull from Jira and Confluence and give me 3 talking points." // INITIATIVE ROLLUP "Give me a plain-English summary of where the [initiative/epic name] initiative stands. What's done, what's in progress, what's blocked, and what's the projected completion?" // DOCUMENT DRAFT — SharePoint hub "Help me draft a one-page FAQ for my SharePoint self-service hub. Topic: How to request a digital wallet enablement for a Synchrony merchant. Audience: internal SyF teams who send me ad hoc requests. Format: short intro, numbered process steps, contact info placeholder."
Scenarios — How to Build "Sub-Agents" in Rooster
What ROVO calls sub-agents are actually Scenarios — routing branches inside ONE agent. Here's the right architecture.
✅ What "sub-agents" are
Scenarios: conditional routing branches inside ONE agent. Different instructions, different knowledge sources, different tools per topic.
⚠️ What they're NOT
Separate agents you can call from Rooster. True agent-to-agent orchestration doesn't exist in Studio yet. Same Rooster, smarter routing.
How to add one
ROVO Studio → Open Rooster → "Add Scenario" (or "Add Sub-agent") → set trigger, instructions, knowledge, tools.
// WHAT "SUB-AGENTS" ACTUALLY ARE IN ROVO STUDIO // ───────────────────────────────────────────── // What ROVO calls "sub-agents" in the Studio UI are actually Scenarios — // conditional routing branches inside ONE agent. Rooster stays one agent. // Each Scenario fires when its trigger matches your prompt. // Different instructions, different knowledge sources, different tools per scenario. // Max 5 tools per scenario for reliable performance (Atlassian's guidance). // ───────────────────────────────────────────── // ROOSTER'S SCENARIO ARCHITECTURE // ───────────────────────────────────────────── // In ROVO Studio → Open Rooster → Add Scenario (or "Add Sub-agent") // Create one scenario per topic area below. // SCENARIO 1: Sprint & Jira // ───────────────────────────────────────────── Trigger: "when the user asks for sprint reports, ticket summaries, Jira data, backlog status, velocity, initiative progress, or anything about what teams are working on" Instructions: You are Rooster in Jira mode. The user wants data from their Jira projects. - Pull sprint status across the requested team(s) using Jira search - Summarize in plain English: what shipped, what's in progress, what's blocked - For accountability decks: one section per team (2-3 sentences + any metrics available) - Flag any issues that have been sitting more than 2 sprints without movement - Format output as structured sections Topher can copy directly into a deck or email Knowledge: [Emerging Payments Jira project] [EHS/UniFi Jira project] Tools: Jira Search, Jira Read // SCENARIO 2: Deck & Document Content // ───────────────────────────────────────────── Trigger: "when the user asks to draft deck content, write a status update, build slide copy, generate document content, or create formatted output for a PowerPoint or Word document" Instructions: You are Rooster in document mode. The user needs structured content they can paste into a real document. - Output in slide-by-slide format when building deck content: "Slide 1 — [title]" then bullet points - Keep each bullet to one line — presentation-ready, not prose - For status updates: lead with the headline, then supporting detail, then any asks - For Bijayta deck: one section per team with ROI signal, sprint summary, key metric, and 1 roadblock max - Offer to regenerate any section in a different format or length - Always end with: "Copy any section above directly into your deck/doc. Want me to rework any section?" Knowledge: [Your personal Confluence page with context] [Emerging Payments Confluence space] Tools: Confluence Search, Jira Search // SCENARIO 3: Meeting & Stakeholder Prep // ───────────────────────────────────────────── Trigger: "when the user is preparing for a meeting, needs talking points, wants to prep for a 1:1 or exec check-in, needs an agenda, or asks what to cover with a specific person" Instructions: You are Rooster in meeting prep mode. - For 1:1s with POs: pull their recent Jira activity (last 2 weeks), surface what moved and what's stalled, suggest 2-3 coaching questions - For exec meetings (Bijayta, Taylor, Mike S): lead with business outcomes, not process details. Frame in ROI, risk, and momentum. - For steering committee: one talking point per team, 30 seconds each, lead with the most interesting data point - Format as a numbered talking points list with time estimates - Keep it to what fits in a 5-minute review before the meeting starts Knowledge: [Your personal Confluence page] [Emerging Payments Jira] [Confluence space] Tools: Jira Search, Confluence Search // SCENARIO 4: PO Coaching // ───────────────────────────────────────────── Trigger: "when the user asks about coaching a PO, documenting performance, preparing quarterly coaching notes, or reviewing a direct report's work patterns" Instructions: You are Rooster in coaching mode. - Pull recent Jira activity for the named PO: stories closed, stories started, stories still open - Identify patterns: Are they closing work consistently? Is backlog health improving? Any recurring blockers? - Suggest 2-3 open-ended coaching questions based on what you see - Format as: Summary paragraph → Data points → Suggested questions - Never include candid performance assessments — keep all output appropriate for company systems Knowledge: [Emerging Payments Jira] [EHS/UniFi Jira] Tools: Jira Search // SCENARIO 5: Synchrony Pay // ───────────────────────────────────────────── Trigger: "when the user asks about Synchrony Pay, the vision deck, the UniFi rebrand, multi-channel checkout, or the moonshot" Instructions: You are Rooster in Synchrony Pay mode. This is Topher's strategic initiative. - Synchrony Pay is the proposed rebrand/repositioning of UniFi as a multi-channel checkout product (in-store + phone + chat) serving dual card, cobrand, AND private label merchants - Today UniFi is private-label ecom only — Synchrony Pay expands the value proposition - Help draft vision deck slides, talking points, and executive framing - Frame in terms of: merchant value, Synchrony differentiation, current capability vs. future state - Keep all content pitch-ready — assume the audience is senior SyF leadership Knowledge: [Your personal Confluence page] [EHS Confluence space] Tools: Confluence Search
Document Output — Workarounds for the Confluence Write Block
IT restricted Confluence write. Here are 5 creative paths to get real documents out of Rooster today.
🟢 Option 1: HTML output (zero approvals)
Ask Rooster for HTML → save as .html → open in browser → Ctrl+P → Save as PDF. Works today. Produces a real formatted document.
🟢 Option 2: OneNote pipeline (already approved)
Rooster generates Markdown → paste into OneNote → format → export to Word or PDF. OneNote is Tier 2 approved and already in your stack.
🟡 Option 3: SyF GPT as the formatter
Rooster retrieves & thinks → raw output to SyF GPT → SyF GPT formats into deck/doc structure. Split the labor: Rooster has context, SyF GPT handles format.
🟡 Option 4: Slide-by-slide copy-paste
Ask Rooster to output in SLIDE N: format → paste each block directly into your SyF PowerPoint template. Fastest path for decks today.
// THE PROBLEM: IT has restricted Rooster from writing to Confluence. // Rooster generates content in chat — but can't push it anywhere. // Here are the workarounds ranked by how much effort they take. // ───────────────────────────────────────────── // OPTION 1: HTML OUTPUT (works today, zero approvals needed) // ───────────────────────────────────────────── // Ask Rooster to output formatted HTML instead of plain text. // Save the HTML as a .html file → open in any browser → print to PDF (Ctrl+P → Save as PDF). // This produces a real formatted document from any Rooster output. Prompt: "Generate this [status report / deck outline / summary] as raw HTML with proper formatting. Include: a header with title and date, styled sections with h2 headings, bullet lists for talking points, and a simple table if there's tabular data. Use inline CSS only (no external stylesheets). I will save this as an .html file and open it in a browser." // ───────────────────────────────────────────── // OPTION 2: ONENOTE PIPELINE (approved stack, works today) // ───────────────────────────────────────────── // Rooster generates content → Topher pastes into OneNote (already Tier 2 approved) → // OneNote has full rich formatting → Export to Word or PDF from OneNote. // OneNote also syncs to Teams/SharePoint automatically. Prompt: "Draft the content for my [document type]. Format it for easy pasting into OneNote: use clear headers (##), bullet points (-), and tables in Markdown format. I will paste this directly into OneNote and format it there." // ───────────────────────────────────────────── // OPTION 3: SYF GPT AS THE FORMATTER // ───────────────────────────────────────────── // Rooster does the retrieval and thinking (Jira data, context) → // Topher copies Rooster's raw output → pastes into SyF GPT with a formatting prompt → // SyF GPT handles final formatting and structure. // Split the work: Rooster knows your context, SyF GPT handles the output format. Prompt to Rooster (step 1): "Pull the current sprint status for all 4 PO teams and give me a raw data dump: each team, their sprint items, statuses, and any blockers. Plain text is fine." Prompt to SyF GPT (step 2 — paste Rooster's output): "Format this into a monthly leadership deck. One section per team. Each section: team name, 3-bullet status summary, one metric, one risk/blocker. Use PowerPoint-friendly formatting (short bullets, no paragraphs)." // ───────────────────────────────────────────── // OPTION 4: SLIDE-BY-SLIDE COPY-PASTE (fastest for decks) // ───────────────────────────────────────────── // Ask Rooster to format output as explicit slide content. // Copy each slide block directly into an existing PowerPoint template. // Topher opens the SyF PowerPoint template, Rooster fills the content. Prompt: "Format this as slide-by-slide content for my monthly accountability deck. For each slide use this exact format: --- SLIDE [N]: [TITLE] [Bullet 1] [Bullet 2] [Bullet 3] SPEAKER NOTE: [one sentence of context for my verbal delivery] --- I will paste each block directly into my PowerPoint template." // ───────────────────────────────────────────── // OPTION 5: WHEN M365 COPILOT ARRIVES (the full solution) // ───────────────────────────────────────────── // M365 Copilot in PowerPoint can generate a complete deck from a prompt or outline. // Workflow once M365 Copilot is enabled: // Rooster → generates outline/content → Topher pastes outline into PowerPoint Copilot → // Copilot generates formatted slides using SyF brand template. // This is the end state. Push IT to prioritize your M365 Copilot request.
Growing Rooster Over Time
Week 1-2: Foundation
- ✓ Create Rooster in Studio (Private)
- ✓ Paste identity
- ✓ Connect Jira projects
- ✓ Create Pages 1-3 in personal space
- ✓ Run privacy tests
- ✓ Run first 5 prompts to see what works
Month 1: Train It
- Add Page 4 (Active Initiatives) after PI Planning
- Start Saved Prompts page from what works
- Connect remaining Confluence spaces
- Run Rooster for one real Bijayta deck cycle
- Tune instructions based on output quality
Month 2+: Automate
- Wire Jira Automation → Rooster for weekly sprint digest
- Set up Monday morning sprint health post to Confluence
- Build PO coaching log prompts into Saved Prompts page
- Consider: publish Rooster to Team once it's polished