Follow-up · Oct 19 callThe ROI case, for your team

The people you're not getting to, found in one pass.

A yes comes down to ROI: people your team wouldn't find on its own, found faster than today. So after the call we ran five OCI searches. 218 director-plus leaders once worked a data center site in a small-market town, and 40 of them have Texas ties. 1,252 distinguished and senior principal engineers at your competitors have a Scholar or GitHub record, often with almost nothing on their profile.

Mock of the connector · every number from a live Crustdata pull · Oct 20, 2026
titles and companies shown, names withheld
370
data center roles Oracle has posted since Aug 1
218
director-plus leaders who've worked a small-market data center site
4,293
distinguished and senior principal engineers at 14 OCI talent competitors
1,252
of those engineers with a Scholar or GitHub record
What I heard

Find the people nobody else is finding, and get to them faster.

It's about finding people that we wouldn't be finding, or it would be very, very time-consuming for us to find on our own.
You get a new search and you're like, where do I start?
There's either nothing on them, or you're having to search patents.
If we can find people who helped start up a data center in the middle of nowhere somewhere else, then they probably will come in and do this.
Just give me the quick hits on why.
What it has to do

Four things it has to get right for Oracle ELR.

VP and above, plus the IC6 and IC7 engineers who earn like VPs, worldwide, with no outside search firms. This is the bar.

01

Start the search for you

Name the competitors, the products and the titles to search, then find the people, in the same pass. That's the step that eats the first hours of every new search.

02

Find who keyword search misses

A distinguished engineer with one line on their profile. A site leader whose small-town build is in their job history, not their headline.

03

Give the quick hits on why

Company, years, the one fact that makes them right. Short enough to paste into the recruiter's notes as is.

04

Keep watching

Execs coming and going across your competitor set, and the people already on your shortlists, every week instead of at the next search.

The ROI case

What you get that profile search and ZoomInfo don't.

Contact data isn't what decides it. Who to call is, and how fast you get to that list. Here's the difference, measured on OCI's market on Oct 20.

People you wouldn't find

Their story is in the history, not the headline.

  • 57of 200 distinguished and senior principal engineers we checked had under 25 words on their profile. Every one has a Scholar record that says what they work on.
  • 160senior leaders worked a data center site in a small-market town, then moved states. It only shows in the dates and places of their past roles.
  • 173of 205 senior power and cooling leaders for data centers work outside Meta, AWS, Google, Microsoft and NVIDIA, at equipment makers, engineering firms and colos.
Getting there faster

Where to look and who fits, in one pass.

Today, per search
Pick competitors→Their products→Guess titles and keywords→Search→Iterate→Write the why

Hours of research before the first call, then more rounds in chat to find what else to look for.

With Crustdata
The kickoff brief→Where to look, who fits, why→Sheet or CRM

One request does both halves. Your analysts spend their hours on the read, not the hunt. All five searches below ran the day after the call.

Fair's fair: ZoomInfo is built around contact data, and that's what Oracle uses it for. Profile search matches what people wrote about themselves today. Crustdata is the step before the call: one record per person across the public web, so you can search what they've done, where and when.

Five searches, run after the call

Five OCI searches, with the numbers.

Ordered by how fast your team could run them. Play 01 is the Abilene search your TA directors described. We show companies and titles here; the names come back in the pull.

01

Data center leaders who've done a small-town site before, and moved for it.

Run it this weekOne pool, every new site

Oracle has posted 370 data center roles since Aug 1, from Abilene and Shackleford County, TX to Saline, MI, Port Washington, WI and Anthony, NM. Senior Director, Data Center Operations alone has been posted for Abilene, Saline and Anthony since August. The people who fit have done this before somewhere else, and the proof is in their job history: the town, the years, the move. 218 director-plus leaders in the US held a data center role in one of 53 small-market campus towns. 160 have moved states since, and 40 have Texas ties. The biggest groups work at Microsoft, AWS, Google, Meta, Vantage, QTS and Switch today.

How it runs
  1. Search the history, not the headline: data center titles at campus towns like Cheyenne, Boardman, Prineville, New Albany and Boydton. person_search on past role title and location
  2. Keep the ones who moved for the work: a site role in one state, a home or current role in another. experience.employment_details.location
  3. Sort for each site: Texas ties for Abilene, Michigan for Saline, New Mexico for Anthony. education.schools.location

A changed location on a profile is easy to miss. A site role in Boardman followed by one in Texas isn't.

→ Search sheetPer siteNew site or req opens
Worked a small-town site, moved since · a sample of the 160Oct 20, 2026
NowThe site roleTie
VantageSr. Director, Site OperationsCritical facilities, Quincy, WA · 2013 to 2020Lives in TX
MicrosoftSr. Director, New Site IntegrationData center manager, West Des Moines, IA · 2010 to 2018Lives in TX
GoogleData Center Operations Site LeadRed Oak, TX now · Meta, Temple, TX beforeLives in TX
Crow HoldingsDirector, Construction ManagementGoogle data center construction, Midlothian, TX · 2018 to 2022Lives in TX
DataBankRegional DirectorLead critical facilities engineer, Lithia Springs, GA · 2020 to 2022Lives in TX
MetaData Center Campus Facilities DirectorNew Albany, OH · AWS area manager in Umatilla, OR beforeWorked in TX
VerrusHead of OperationsAWS regional manager, Boardman, OR · 2014 to 2016Worked in TX
IRENVP, Next Gen Data Center ArchitectureAWS, Boardman, OR · 2019 to 2022 · a Facebook build in New Albany, OH beforeLives in TX
The 53-town list covers small-market hyperscale campuses. Your team can add or drop any town. Names come back in the pull.
02

IC6 and IC7 engineers with nothing on their profile, found by their papers and code.

The hard ones

These are the seats that pay like a VP and leave the faintest trail. At AWS, Google, Microsoft, Meta, NVIDIA, CoreWeave and eight more companies OCI hires against, we count 4,293 distinguished, fellow, senior principal and partner-level engineers. 1,252 have a Google Scholar or GitHub record joined to their profile. When we checked 200 of the Scholar ones, 57 had under 25 words on their profile and current role, and 112 had an h-index of 20 or more.

How it runs
  1. Pick the seat and the companies: distinguished, fellow and senior principal titles at the competitors you name. person_search
  2. Read what they publish and build: research interests, papers and h-index, plus repos and languages. research_profiles dev_platform_profiles
  3. Rank on the work, with a one-line why: "NVIDIA, distinguished engineer, AI networking and ML for systems, 94 papers."

Patents aren't a field we filter on today. Papers and code are, and for these seats they're often the clearest signal there is.

→ Shortlist with a whyPer searchNew req
Thin profile, deep record · from the 200 we checkedOct 20, 2026
Company · titleWordsPapersWorks on
AWSSenior Principal ScientistAutomated verification, model checking8377Automated verification, model checking
NVIDIADistinguished EngineerAI hardware and software co-design, AI networking094AI hardware and software co-design, AI networking
GoogleDistinguished EngineerNetworked and distributed systems042Networked and distributed systems
MicrosoftPartner ArchitectComputer architecture, HPC035Computer architecture, HPC
AWSSenior Principal EngineerNetworking11124Networking
NVIDIADistinguished EngineerMachine learning, systems5131Machine learning, systems
MetaDistinguished ScientistComputer graphics0110Computer graphics
Words counted across the profile summary and current role description. Names come back in the pull.
03

Power and cooling leaders, and the competitors you didn't know you had.

New territory

OCI now posts roles like Lead Principal Electrical Engineer for data centers and Senior Director, Power Systems Engineering. We found 205 senior power, cooling and mechanical leaders for data centers in the US. Only 32 work at Meta, AWS, Google, Microsoft or NVIDIA. Most of the rest sit at equipment makers like Schneider Electric, Vertiv, Johnson Controls and Trane, engineering firms like Jacobs and AECOM, and colos like Equinix and EdgeConneX. That list is the "where do I start" step, done from the data.

How it runs
  1. Describe the seat in plain words: data center power, cooling, thermal and MEP leadership. person_search on current title
  2. Let the data name the competitors: count where those people work today, by company.
  3. Keep the new names on your competitor list, so they feed the next search and the weekly watch in play 05.

You don't need to know Vertiv's org chart to search it. The map tells you where to look first.

→ Competitor mapPer functionEach new search
Where 205 senior power and cooling leaders work · top employersOct 20, 2026
GroupCompanies · profiles
Hyperscalers and chipsMeta 8 · AWS 7 · Google 6 · NVIDIA 6 · Microsoft 5
Power and cooling makersSchneider Electric 4 · Vertiv 3 · Johnson Controls 3 · Trane 2 · Eaton 2 · Carrier 2 · Siemens 2 · Chemours 2
Engineering firmsJacobs 4 · AECOM 2 · Salas O'Brien · DLB Associates
Colos and neocloudsCsquare 4 · Equinix · EdgeConneX · Fluidstack
Counts are profiles in our index with a data center power, cooling, thermal or MEP title at lead level or above.
04

Corporate searches with a Nashville tie, ready before the kickoff.

Only Oracle can run this one

Oracle's headquarters is moving to Nashville, so corporate searches now carry a location question too. Take investor relations: 2,103 director-plus IR leaders in the US came up through banking or equity research, and 316 of them work in IR at tech companies today. Across all 2,103, 56 have Tennessee ties: they live there, worked there or studied there. That's the 25 profiles and the 25 target companies, in hand an hour before the kickoff instead of after hours of research.

How it runs
  1. Brief it in plain words: an IR seat today, banking or equity research before. person_search on current and past roles
  2. Add the tie: a home, a past role or a school in Tennessee. basic_profile.location education.schools
  3. Bring the target list too: where the pool works today, ranked, so the kickoff starts from companies, not guesses.

Finance isn't the hard search. The Nashville tie is what makes it Oracle's.

→ Kickoff packPer searchKickoff booked
IR at tech companies, from the sell side · a sample of the 316Oct 20, 2026
NowBefore
FortinetVP, Investor RelationsJ.P. Morgan · equity analyst, IT hardware
Micron TechnologyCorporate VP, IR and TreasuryCredit Suisse · director, equity research
CloudflareVP, Strategic Finance, Treasury and IRCredit Suisse · managing director, equity research
Astera LabsSVP, Finance and IRB. Riley · equity research, semiconductors
BILLVP, IR and Capital MarketsNeedham · senior research analyst, software
AvalaraVP, Investor RelationsCitizens JMP · equity research, software
AsanaHead of Investor RelationsUBS · equity research
56 of the 2,103 have Tennessee ties. Names come back in the pull.
05

Watch the execs coming and going, and the people you've already found.

Runs every week

Tracking competitors over time is half the job. Since Jul 20 we count 31 VP-level and distinguished hires into your competitor set from outside, and 9 of them came from another company on the same list. A weekly watch sends those moves to the team, and flags it when someone on a past shortlist changes jobs, titles or cities.

How it runs
  1. Set the list once: your competitors, plus everyone on your recent shortlists. watch_create
  2. It checks every week for new seats, title changes, moves and new cities.
  3. The moves arrive as a weekly digest by email or Slack, and later in your CRM.

You hear about the move the week it shows up, not when the next search starts.

→ Weekly digestWeeklyA seat changes
New seats at your competitor set since Jul 20 · from outsideOct 20, 2026
JoinedFromWhen
CoreWeaveSVP, EngineeringGoogle · VP, Engineering, Distributed CloudOct
GoogleDistinguished EngineerGlean · chief security officerSep
GoogleVP of EngineeringJohnson & Johnson · SVP and CIOSep
NVIDIADistinguished EngineerHPE · senior distinguished technologistAug
NVIDIADistinguished Research ScientistMeta · senior staff research scientistAug
NVIDIADistinguished EngineerMicrosoft · principal security engineerAug
NebiusDistinguished EngineerPliops · VP, technologyAug
Internal promotions and advisory roles left out. Names come back in the watch.
Where it plugs in

Your enterprise ChatGPT now. Your new CRM next.

Your team searches in plain language where it already works. The same records that fill the sheet today can write into the exec search CRM you're building, so recruiters don't re-type them.

Already yours

Enterprise ChatGPT

Oracle's own instance. Your analysts ask in plain language, the way they brief a search today.

New

Crustdata connector + skill

Adds 1B+ people and 60M+ companies to it, plus a skill built around your searches: talent maps, sourcing, competitor tracking.

Today

The search sheet

Ranked lists with the quick hits on why, ready for recruiters to call from.

Next year

Your exec search CRM

Build it with our API in mind and the records flow straight in, refreshed as people move.

How to prove it

Run three searches side by side with what you use today.

Score the outputs, not the pitch. Each ask comes with our number from Oct 20.

Ask 01Coverage

The Abilene search

Count senior leaders who've worked a small-market data center site and moved for it. Then count how many have Texas ties.

Ours: 218 who worked one, 160 who moved, 40 with Texas ties.
Ask 02Depth

IC7s with a thin profile

Pick a competitor and a seat. Count the distinguished engineers you can describe in one line from their work, not their headline.

Ours: 57 of 200 checked had under 25 words on their profile, and all 200 have a paper trail.
Ask 03Time

Kickoff notes to a ranked list

Start the clock when the kickoff notes are in. Stop it at a ranked list with a why on every name.

Ours: all five searches on this page ran the day after the call.
Straight answers

The questions from the call.

Q1Is there an interface, or do we connect our enterprise ChatGPT?

Connect it. Crustdata adds as a connector in your enterprise ChatGPT, and your team searches there in plain language. We build a skill with you for your searches: the guardrails and logic for talent mapping, sourcing and competitor tracking, tuned to how your analysts work.

Q2Where does it show where someone appears across the web?

Each person comes back as one record: work history with dates and locations, education, Scholar publications, GitHub, posts and news, all resolved to the same person. The summary gives the quick hits, and the sources are there for your team to verify.

Q3Is it the public profile data plus other sources?

Yes. Public professional profiles plus GitHub, Google Scholar, personal sites, posts and news, mapped to one record per person and kept fresh. That covers 1B+ people and 60M+ companies.

Q4Is there a contact information piece?

Yes: work and personal emails and phone numbers, added when the team is ready to reach out. It's there when you want it. The search is what this page is about.

Q5Can it search patents?

Not as a field today. Publications, code, personal sites and work history are, which is how play 02 finds the engineers with nothing on their profile.

What happens next

Easy to try. Let's make the buy easy too.

Everything below came out of the call. No dates we didn't agree to.

Pricing comes with the next conversation, sized to your team and your searches.

The short version

The people you're not getting to, found in one pass, with the why.

Where to look and who fits, in the same request, in the ChatGPT your team already uses. Five OCI searches above show what that looks like on your market.