Executive Summary
30-day post-close organizational assessment
Est. 12-Month Turnover Cost Risk
$670K – $1.06M
High Attrition Risk
3
Critical 90-day window
Critical Manager Flags
2
Require immediate intervention
Integration Friction
5%
Grid↔AI single-bridge risk
Key Findings
Critical
AI Compute Tri-Lead Fragmentation
AI Compute runs three teams under three directors (Zara Karim · FMT, Priya Venkatesan · Inference & Edge, Wren Collier · Observability) with materially different work cultures — heads-down research, hyperscaler SLA cadence, and on-call incident command. The division reads as three subcultures, not one. Cultural strength: 2.4/10.
Critical
Grid↔AI Single-Bridge Risk
The vertically-integrated thesis rests on one informal tie: Tamsin O'Neill (Renewable Offtake) → Priya Venkatesan (Inference & Edge). 1 edge out of 22. No redundancy. Either departure decouples 1.2 GW signed PPA portfolio from inference-side demand planning.
High
Key Person Risk · Cassian Barlow
Engineer, Edge Deployment (Cassian Barlow) — junior on paper, but Omar Nassif routes nearly all production deployments through him. Quiet broker in the I&E stack. Replacement cost: $194K. At-risk window: 90 days.
High
Sustainability Pipeline Single-IC Failure
Mateo Cruz (Observability Engineer) is the sole formal tie between AI Observability and Net Zero Ops via Leena Nair. HIGH flight risk. Departure collapses the only AI-to-sustainability data pipeline. Replacement cost: $258K.
Recommended Actions
1
AI Compute Cultural Carrier Protection
Urgent
2
Grid↔AI Integration Redundancy Build-Out
High
3
Key Person Retention Program · Cassian Barlow & Jules Moreau
Urgent
4
Director Coaching · Wren Collier & Tamsin O'Neill
High
5
Tri-Division Vocabulary Mapping & Communications Cadence
Strategic
Trusted Advisors
10
Network centrality analysis
Top Centrality
4
Zara Karim · AI Compute
Network Density
38%
Leadership-tier cluster
Top 10 by Combined Network Degree
| Rank | Name | Division | Degree | Status |
|---|
Avg Trust Rating
77%
Across leadership tier
High Performers
6
Composite score 75%+
Intervention Needed
2
Below 60% composite
Manager Composite Scores
Detailed Manager Assessment · SVP + Team Leads
| Manager | Division | Trust % | Advice % | Comm % | Composite % | Status |
|---|
Critical Gaps
1
Grid↔AI · single-bridge
Highest Traffic
Executive
Cross-division arbitration
Division Communication Density Matrix · %
| Executive | AI Compute | Grid Power | Sustainable Infra | |
|---|---|---|---|---|
| Executive | 88 | 55 | 48 | 42 |
| AI Compute | 55 | 78 | ⚠ 9 | 22 |
| Grid Power | 48 | ⚠ 9 | 82 | 28 |
| Sustainable Infra | 42 | 22 | 28 | 65 |
Communication Load by Division
High Risk · 90D
3
Immediate intervention
Moderate Risk · 180D
2
Monitoring & engagement
Est. Replacement Cost
$1.39M
If all 5 depart unmanaged
Cassian Barlow
Engineer, Edge Deployment · AI Compute
High Risk
Junior on paper, but Omar Nassif routes nearly all production deployments through him. Quiet broker in the Inference & Edge stack. 10-year tenure with no title progression — promotion-cycle frustration is the dominant departure signal. Replacement cost: $194K.
Trust Percentile34%
Cultural Fit40%
Replacement cost · $194,000 · Production deployment continuity at risk
Jules Moreau
Mechanical Engineer, Cooling · Grid Power
High Risk
Title says mechanical engineer, but rack-level designs are the dependency for both high-density cooling AND renewable offtake load profiles. Higher informal centrality than his level suggests — and zero formal cross-division recognition. Replacement cost: $217K.
Trust Percentile38%
Cultural Fit44%
Replacement cost · $217,000 · Cooling–PPA load integration dependency
Mateo Cruz
Observability Engineer · AI Compute
High Risk
Sole formal tie between AI Observability and Net Zero Ops via Leena Nair. Departure collapses the only AI-to-sustainability data pipeline. Single-point-of-failure for scope 2 emissions reporting on inference workloads. Replacement cost: $258K.
Trust Percentile32%
Cultural Fit41%
Replacement cost · $258,000 · Sole AI↔Sustainability data bridge
Wren Collier
Director, AI Observability · AI Compute
Moderate Risk
Two unsolicited recruiter conversations in Q1. Compensation at 88th percentile but title is a lag — Director-level pay with VP-level scope (reliability incident command, on-call governance, SLO/error budget policy). Replacement cost: $321K.
Trust Percentile56%
Cultural Fit58%
Replacement cost · $321,000 · AI Observability leadership at risk
Tamsin O'Neill
Director, Renewable Offtake · Grid Power
Moderate Risk
Sole network bridge between Grid Power and AI Compute via Priya Venkatesan. Carries the only PPA-to-inference relationship. 3-year tenure, just enough to be poachable. 1.2 GW signed PPA portfolio walks with her. Replacement cost: $402K.
Trust Percentile60%
Cultural Fit62%
Replacement cost · $402,000 · Sole Grid↔AI integration relationship
Read the White Paper · What Most (PE)ople Get Wrong About Culture→
Read the white paper · "What Most (PE)ople Get Wrong About Culture" →
How to Read the Cultural Data
Cultural strength — not cultural content — predicts post-close integration success.
How to Read the Cultural Data
Cultural strength — not cultural content — predicts post-close integration success.
Most culture assessments collapse three distinct questions into a single score. G-Engine separates them:
Cultural Content
What is the culture? The vocabulary employees use to describe how work gets done. Easiest to observe; least predictive of performance.
Cultural Fit
Who is aligned to it? The linguistic distance between each employee and the modal organizational vocabulary. The leading indicator of voluntary departure 3–6 months in advance.
Cultural Strength
How uniformly is the culture held? The variance of fit across the workforce. Determines whether the org responds with coordination or fragmentation under integration pressure.
AI Compute · Tri-Lead Fragmentation
Low-Strength · Carrier-Dependent
AI Compute reads as three subcultures under three directors — Zara Karim's training research culture (heads-down, depth-first, "the on-call is brutal but the compute is unreal"), Priya Venkatesan's customer-facing inference cadence (hyperscaler SLA urgency, "latency is the only currency that matters"), and Wren Collier's reliability practice (on-call rhythm, incident command, "error budget or it didn't happen"). No shared vocabulary across the three. Under integration pressure the division will respond like three teams, not one. Cultural-carrier protection priority: Wren is two recruiter conversations away from leaving.
Grid Power · The Execution Anchor
High-Strength · Aligned With Thesis
Grid Power registers as the strongest-cohesion division. Long tenures (Darius 9y, Chen 10y, Anika 11y), LOW flight risk across the leadership tier, and a shared engineering vocabulary that connects high-density cooling, grid-scale interconnect, and PPA structuring ("megawatts before models," "the substation is the moat"). This is the moat that justifies the vertically-integrated thesis. The structural risk: Grid Power talks within itself but is tied to AI Compute by exactly one informal relationship — Tamsin O'Neill → Priya Venkatesan. A strong division that cannot bridge is still a coordination liability.
Cultural Strength by Division
Strength = inverse variance of individual cultural fit within each division. Higher = more uniformly held vocabulary; lower = fragmentation. Survey free-text scaffold for demo — live engagements populate from client respondents.
Cultural Content · Organizational Vocabulary
The dominant words employees use to describe Heinlein. This is what the culture is — useful context, but not by itself predictive of integration outcomes (see strength, above). Scaffold drawn from peer-set discourse (CoreWeave, Crusoe, Bloom Energy, Watershed, Generate Capital) — replaced by client free-text in live engagements.
Compute
Throughput
Megawatts
On-call
Reliability
Latency
Sustainability
Hyperscaler
Burnout
PPA
Cooling
Grid
Net-Zero
Inference
Substations
SLO
Single-Point
Carbon
Edge
Recruiter-Pings
Backlog
Tokens
Vendor-Lock
Ramp
01
TimelineWithin 14 days
ObservationAI Compute runs as three subcultures under Zara Karim (FMT), Priya Venkatesan (I&E), and Wren Collier (AI Observability). Composite cultural strength: 2.4/10. Wren has had two unsolicited recruiter conversations in Q1; compensation at 88th percentile but title is a lag.
Research BasisThree-director divisions with no shared cultural vocabulary respond to integration pressure with fragmentation, not coordination (variance-of-fit measure, Aven prior portfolios). Mid-tenure directors (3–4y) with comp/title mismatch index in the top quartile of pre-departure signal.
Specific RiskWren's departure forces collapse of two AI Compute teams into one; loss of incident command leadership; reliability incidents become uncoordinated. Replacement cost: $321K. Title elevation to VP-equivalent within 30 days, paired with explicit retention bonus on hold-period vest, reduces 90-day departure probability by ~45%.
02
TimelineCross-tie by day 21; 8-week relationship build-out plan
ObservationSingle informal tie connects Grid Power to AI Compute: Tamsin O'Neill (Director, Renewable Offtake) → Priya Venkatesan (Director, Inference & Edge). 1 edge out of 22. The PPA-to-inference business model rests on one relationship.
Research BasisNetwork theory: a graph with one bridge between two clusters is a structural-hole-of-one — maximum brokerage value, maximum bridge-failure risk (Burt 2004; replicated in PE-portfolio integration data). Tamsin is MEDIUM flight risk; comp 80th percentile, title lag.
Specific RiskEither departure decouples 1.2 GW signed PPA portfolio from inference-side demand planning. Cascading impact across grid-scale interconnect filings, hyperscaler SLAs. Build at least two more Grid↔AI ties (Gabriel Flores → Omar Nassif; Darius Khoury → Wren Collier) before 90 days.
03
TimelineCustomized engagement plan by week 2; ongoing through 90-day window
ObservationCassian Barlow (Engineer, Edge Deployment) is titled junior but routes nearly all production deployments — quiet broker in I&E. Jules Moreau (Mechanical Engineer, Cooling) holds rack-level designs that the cooling and PPA load-profile teams both depend on. Both HIGH flight risk. Mateo Cruz (Observability Engineer) is the third HIGH-risk hidden broker — sole AI Obs↔Net Zero tie.
Research BasisAll three rank in top-10 by combined network degree despite IC-level titles. Burt-style structural-hole analysis: their downstream-affected counts in the Digital Twin simulator exceed several directors above them. 10-year tenure (Cassian) with no title progression indexes high in voluntary-departure literature.
Specific RiskCombined replacement cost $669K. Production deployment continuity (Cassian), cooling-PPA integration (Jules), and AI-sustainability data pipeline (Mateo) are all single-IC dependencies. 90-day window.
04
TimelineAssessment by week 3; coaching engagement begins week 4
ObservationTwo directors flag for coaching, both MEDIUM flight risk: Wren Collier (composite 58%, comp lag) and Tamsin O'Neill (composite 59%, sole-bridge stress). These are the two single-points-of-failure flagged in Action 02 from the manager-side.
Research BasisMid-tenure directors (3–4y) with high informal centrality and recruiter activity index in the top quartile of pre-departure signal. Coaching plus comp/title alignment within 30 days reduces 90-day departure rate by 40–50% in PE-portfolio benchmarks (Aven prior assessments).
Specific RiskWren's departure collapses AI Observability. Tamsin's departure decouples Grid↔AI. Combined upstream replacement cost $723K, not counting brokerage-loss effects modeled in the Digital Twin.
05
TimelineComprehensive plan by day 30; 180-day rolling communication cadence
ObservationThree divisions with three distinct vocabularies — AI Compute (research/SLA/on-call), Grid Power (interconnect/PPA/thermal), Sustainable Infrastructure (carbon/scope-reporting). Overall cultural fit 5.4/10 (Grid 7.2, Sustainable 6.1, AI Compute 2.4). No shared operational language across them yet.
Research BasisCultural fit predicts voluntary departure 3–6 months in advance; cultural strength (variance of fit) determines whether the org responds to integration pressure with coordination or fragmentation. Tri-pillar AI infra companies historically resolve to either (a) one dominant vocabulary or (b) sustained fragmentation — no stable middle.
Specific RiskSponsor (Aven Capital) thesis is vertically-integrated — requires coordination across all three pillars. Without deliberate vocabulary unification, the platform reads to talent as three companies that share a CFO. Each 0.5-point cultural fit decline correlates with 2–3% additional at-risk population.
What-If · Drop A Person Out
The formal org chart says who reports to whom. The informal network shows who the work actually flows through.
Select an employee. The left panel shows their position on the formal hierarchy. The right panel shows the real information-seeking network among the leadership tier. Removing them recomputes structural-hole exposure, brokered ties severed, replacement cost, time to refill, and the downstream employees who lose access. Use this to surface quiet brokers whose departure looks low-impact on the org chart but high-impact on the actual network.
Formal Org Chart
Behavioral Network Analysis
Select an employee above to simulate their departure and see the network impact.