engineering-org-chart

Engineering Org Chart — Skill Instructions

What this skill does

Maps the full engineering hierarchy at a target company using two data layers:

• Layer 1 (Prospects): reo_list_prospects filtered to engineering tech_functions — authoritative people list
• Layer 2 (Signals): reo_get_account_activities — who has actually touched your product and when
• Output: Interactive HTML tree chart, nodes coloured by signal status, grouped by sub-team

⚠️ CRITICAL DATA SOURCING RULE: The people list comes ENTIRELY from reo_list_prospects. Do NOT derive the people list from activity signals. Signals tell you WHO IS ACTIVE, not WHO EXISTS. Using activity data to discover people will surface whoever uses the product (often the customer's sales/GTM team), NOT the engineering org you want to map.

---

CONFIG

SIGNAL_ACTIVE_DAYS: 30

Node is GREEN 🟢 if they have an activity signal within this many days.

Node is AMBER 🟡 if a signal exists but is older than SIGNAL_ACTIVE_DAYS.

Node is GREY ⚫ if no signals at all.

SIGNAL_LOOKBACK_DAYS: 90

SHOW_LINKEDIN_LINKS: true

---

Step 1 — Resolve Company Domain

If the user gives a company name, infer the domain (e.g. "Notion" → "notion.so"). If ambiguous, ask the user to confirm.

---

Step 2 — Fetch Team Headcount (reo_get_tech_function_job_counts)

Call reo_get_tech_function_job_counts with the domain. Returns headcount by tech_function. Use to:

• Understand team composition before querying people
• Show totals in the chart header
• Skip sub-teams with count = 0 (no call needed for empty teams)

The valid tech_function values from this API are:

• "Engineering Leadership"
• "Artificial Intelligence & Machine Learning"
• "Backend Engineering"
• "Cloud & Infrastructure"
• "Full Stack"
• "Frontend Engineering"
• "DevOps Platform & Reliability"
• "Security & Compliance"
• "QA Automation & Testing"
• "Infrastructure & IT Operations"
• "Mobile Engineering"

---

Step 3 — Fetch All Engineering People (reo_list_prospects)

⚠️ THIS IS THE ONLY SOURCE OF PEOPLE. Not activity data. Not open data profiles.

Call reo_list_prospects with:

• domain: the resolved company domain
• tech_function: ALL engineering functions from Step 2 that have count > 0 (see list above)
• page_size: 50

Check pagination.total_pages in the response. If > 1, make additional calls with page_number: 2, page_number: 3, etc. until all pages are fetched.

Do not filter by seniority — fetch all levels so you can build the full hierarchy.

For each person capture:

• id — LinkedIn public_id (used for URL construction and signal matching)
• full_name
• designation — job title
• seniority — reo seniority label (use directly, don't re-infer from title)
• tech_function — array, used to assign sub-team
• linkedin — full LinkedIn URL
• is_likely_to_engage — show ⚡ badge if true
• current_position_start_date
• state — US state or equivalent region
• country — country name

---

Step 4 — Classify: Hierarchy Level + Sub-Team

4a. Hierarchy Level — use reo seniority field directly

reo seniority	Level	Display label
C-Level	0	👑 C-Suite
VP	1	🏆 VP / Head
Director	2	🎯 Director / Principal
Manager	3	🧩 Manager / Staff
Senior	4	⭐ Senior / Lead
Mid	5	💻 Mid-level IC
Junior	6	🌱 Junior
null / unknown	5	💻 IC

4b. Sub-Team — map directly from tech_function array

reo tech_function	Sub-Team
Engineering Leadership	🏛️ Engineering Leadership
Artificial Intelligence & Machine Learning	🤖 AI / ML
Backend Engineering	🖥️ Backend
Cloud & Infrastructure	☁️ Cloud & Infra
Full Stack	🔧 Full Stack
Frontend Engineering	🎨 Frontend
Mobile Engineering	📱 Mobile
DevOps Platform & Reliability	⚙️ DevOps / Platform
Security & Compliance	🔒 Security
QA Automation & Testing	🧪 QA / Testing
Infrastructure & IT Operations	🛠️ Infra & IT

If tech_function is null/empty, fall back to title keywords (see Appendix at bottom).

4c. Location Group — derive from state + country fields

Build a location_group label for each person:

if country == "United States" and state is not null → location_group = state  (e.g. "California")
elif country is not null                            → location_group = country (e.g. "India")
else                                                → location_group = "Unknown"

After processing all people, consolidate US states with < 2 people into "Other US" to keep the chart readable if there are many distinct states.

---

Step 5 — Fetch Activity Signals (reo_get_account_activities)

Call reo_get_account_activities with the account ID (resolve domain → UUID via account lookup first if needed).

For each activity event extract: person identifier, activity_type, timestamp.

Build signal map: prospect id → signal status

For each person from Step 3:
  If matched to activity AND last signal within SIGNAL_ACTIVE_DAYS  → 🟢 "active"
  If matched to activity AND last signal older than SIGNAL_ACTIVE_DAYS → 🟡 "stale"
  No match → ⚫ "none"
  Store: last_seen date, total event count, top pages visited

If activity data returns people NOT in the engineering prospects list (e.g. sales/GTM users):

• Include them in the signal map but do NOT add them to org chart nodes
• Collect their names/titles for a footnote below the chart

---

Step 6 — Render the HTML Org Chart (show_widget)

Use show_widget. Always render as HTML — never output a markdown table.

Color palette

• Page bg: #111827; Panel bg: #1f2937; Node bg: #374151
• Active border/dot: #4ade80; Stale: #f59e0b; None: #4b5563 / #6b7280
• Text primary: #f9fafb; secondary: #9ca3af; muted: #6b7280
• Sub-team header: #1e3a5f bg; Connector lines: #374151
• Accent: #60a5fa blue, #a78bfa purple

Layout — Two View Modes with Toggle

Add a toggle bar in the header with two buttons: [ By Team ] [ By Location ] Default view is By Team. Clicking the toggle switches the swimlane dimension without reloading — use JavaScript to show/hide the two pre-rendered views.

View A — By Team (default)

┌─────────────────────────────────────────────────────────┐
│ HEADER: [Company] Engineering Org                        │
│ [N] engineers · 🟢 [N] active · 🟡 [N] stale · ⚫ [N]  │
│ [ By Team ▼ ]  [ By Location ]    Legend: 🟢 🟡 ⚫      │
├─────────────────────────────────────────────────────────┤
│         👑 C-Suite & 🏆 VP Eng  (full-width row)         │
│         [NODE]  [NODE]  ...                              │
├──────────┬──────────┬─────────┬──────────┬──────────────┤
│ 🤖 AI/ML │🖥️ Backend│☁️ Cloud │🎨 Frontend│⚙️ DevOps ... │
│ [L2–L6 nodes sorted by seniority within each column]    │
└──────────┴──────────┴─────────┴──────────┴──────────────┘

View B — By Location

┌─────────────────────────────────────────────────────────┐
│ HEADER (same)                                            │
│ [ By Team ]  [ By Location ▼ ]    Legend: 🟢 🟡 ⚫      │
├────────────────┬───────────────┬──────────┬─────────────┤
│ 📍 California  │ 📍 New York   │ 📍 India │ 📍 Other... │
│ [N people]     │ [N people]    │ [N]      │             │
│                │               │          │             │
│ Grouped by     │ Grouped by    │ Grouped  │             │
│ sub-team       │ sub-team      │ by team  │             │
│ within col     │ within col    │          │             │
│                │               │          │             │
│ [NODE cards,   │ [NODE cards]  │ [NODES]  │             │
│  sorted by     │               │          │             │
│  seniority]    │               │          │             │
└────────────────┴───────────────┴──────────┴─────────────┘

In the By Location view:

• Each column = one location_group (state for US, country for international)
• Show the location flag emoji if country is known (🇺🇸 🇬🇧 🇮🇳 etc.)
• Within each column, group nodes by sub-team label (small grey sub-header), then by seniority
• Show "📍 [Location] · [N] engineers" as the column header
• Sort columns by headcount descending (largest location first)

Node anatomy

Each card has:

• Coloured left border (green / amber / grey)
• Initials avatar (bg matches signal colour)
• Full name (links to LinkedIn if SHOW_LINKEDIN_LINKS = true)
• Job title (11px muted)
• Signal line: dot + "Last seen [date]" or "No signals"
• ⚡ badge if is_likely_to_engage = true

Interactivity

• Click node → tooltip: pages visited, activity types, event count, first/last seen
• Sub-team column headers → collapse/expand
• Hover → lift shadow
• If sub-team has > 8 ICs: show first 5 + "Show +N more" chip

Non-engineering activity note (if any)

If Step 5 found active users not in the engineering org:

"⚠️ [N] non-engineering users also active: [names / titles]. Likely internal Reo product users."

🧠 What this tells us

2–3 sentences:

• Which sub-team has the most signal activity
• Which seniority level is engaged (ICs vs leadership)
• Notable gaps (e.g. "No signal from Engineering Leadership — likely bottom-up evaluation")

🎯 Suggested next steps

3 prioritised actions tied to specific people or gaps:

• 🔴 Urgent | 🟡 This month | ⚪ When ready

---

Edge Cases

Situation	Handling
No engineers from reo_list_prospects	Empty state: "No engineering profiles found for [domain]." Show headcount from Step 2.
No signals	Full chart with all ⚫ nodes. Banner: "No product activity — use as prospecting map."
Domain can't be resolved	Ask user to confirm domain.
Large team (50+ engineers)	Collapse ICs by default with "Show all" per sub-team.
Only 1 active engineer	Banner: "⚠️ Single-threaded — high stall risk."
null tech_function	Fall back to title keyword sub-team matching (Appendix).

---

Appendix — Title Keyword Fallback for Sub-Team (when tech_function is null)

Sub-Team	Title keywords
🤖 AI / ML	ML Engineer, Machine Learning, AI Engineer, Data Scientist, NLP, LLM, MLOps
🖥️ Backend	Backend Engineer, API Engineer, Server-side, Golang, Python Eng
☁️ Cloud & Infra	Infrastructure, Cloud Engineer, SRE, Site Reliability
⚙️ DevOps	DevOps, Platform Engineer, Platform Reliability
🔒 Security	Security Engineer, AppSec, InfoSec, CISO
🎨 Frontend	Frontend, UI Engineer, React, Vue, Angular
📱 Mobile	iOS, Android, Mobile, React Native
🔧 Full Stack	Full Stack, Fullstack
🏛️ Engineering Leadership	CTO, VP Eng, Director of Eng, Head of Eng, Engineering Manager