Employer leasing, development, and named site decisions
No probability on Digital Realty occupancy, rents, bookings, or named campuses.
Disclosures · Forecast desk
Conflict register.
I am testing whether a transparent, public-data forecasting process can add signal on a bounded set of AI-infrastructure questions. Every forecast is timestamped and scored. The track record is still early, and I am not claiming a demonstrated forecasting edge.
Employment
D. Brian Letort, Ph.D. is Head of Corporate Data Office & AI at Digital Realty. Register version 2026-09-06. Employer legal and communications written sign-off is requested not received. The public ledger operates as an experiment under this register and does not scale market-moving claims until that status changes.
Dual hat
- Magellan
- Private evidence and denominator engine. Does not emit public probabilities. Hypotheses and scenarios stay internal.
- This site
- Signed public forecast ledger. Every published number must remain verifiable from a public URL if Magellan disappeared.
- Public data only
- Yes. Every number is verifiable from a public URL.
Recusal set
The questions this desk will not price, regardless of how informative they would be.
Named competitor bookings
No probability on a named colocation or hyperscale landlord's contracted MW or named customer win.
Customer pipeline and MNPI
No use of non-public customer, Salesforce, or internal CI predictors.
Employer ticker
No directional call on DLR or DXM securities.
Allowed public inputs
- Public interconnection queues, permits, and moratoria
- Issuer filings and earnings transcripts
- Cited third-party research used as inputs, not as hidden authority
- Industry-composite conversion rates (announced MW vs energized MW)