Get an API key
EN中文

Card 01Jobs

Texas, not Virginia, has the most data center job postings

Open US postings with "data center" in the title, by state, employer and role, plus the electricians and construction leads whose postings name data centers, on the Metix AI Platform on September 30, 2026.

1,148

open data center postings in Texas, 1.9× Virginia's 604.

  1. Texas1,148
  2. Virginia604
  3. California457
  4. Ohio351
  5. Georgia320

Of 6,721 open US postings with "data center" in the title.

Texas posts 1.9 times as many data center jobs as Virginia

Open US postings with "data center" in the title, by state, drawn as a share of all 6,721

  1. Texas1,148 · 17.1%
  2. Virginia604 · 9.0%
  3. California457 · 6.8%
  4. Ohio351 · 5.2%
  5. Georgia320 · 4.8%
  6. North Carolina224 · 3.3%
  7. Oregon222 · 3.3%
  8. Washington214 · 3.2%
  9. Arizona180 · 2.7%
  10. Indiana156 · 2.3%
  11. New York153 · 2.3%
  12. Colorado145 · 2.2%
  13. Every other state2,547 · 37.9%

What it shows

Texas has 1,148, 17.1%; Virginia, the market the industry is best known for, is second with 604. California, Ohio and Georgia follow.

The national total is 30 state counts plus one count for every other state; 986 are in other states or give no state.

Method and limits

A role listed in several cities or reposted counts more than once, so these numbers measure hiring activity, not seats.

Source: Metix AI Platform, jobs, 2026-09-30.

Querydata-center-titles.json
POST /v1/jobs/query · queries/data-center-titles.json
{  "where": {    "all": [      {        "field": "title",        "match": "data center"      },      {        "field": "location.country",        "eq": "United States"      }    ]  },  "size": 1}

Amazon posts nearly a fifth, but Texas's lead is not one employer

Each employer's share of all 6,721 data center postings

  1. Amazon1,272 · 18.9%
  2. Oracle257 · 3.8%
  3. Meta109 · 1.6%
  4. Google97 · 1.4%
  5. QTS Data Centers32 · 0.5%
  6. Serverfarm31 · 0.5%
  7. CoreWeave26 · 0.4%
  8. Equinix15 · 0.2%
  9. Switch10 · 0.2%
  10. Microsoft8 · 0.1%

What it shows

Amazon, which posts under two names, has 1,272, 18.9%. But only 191 of them are in Texas, against 279 in Virginia. Without Amazon, Texas has 957 and Virginia 325: the gap widens.

Microsoft has only 8, which says where it advertises, not how much it builds.

Method and limits

Employers are matched on the exact names the records use. Postings under another spelling, or through a staffing agency, are not in the employer's row, so these are floors.

Source: Metix AI Platform, jobs, 2026-09-30.

Querydata-center-titles.json
POST /v1/jobs/query · queries/data-center-titles.json
{  "where": {    "all": [      {        "field": "title",        "match": "data center"      },      {        "field": "location.country",        "eq": "United States"      }    ]  },  "size": 1}

Technicians and managers lead; one in ten is a construction role

The word the title uses for the job, as a share of all data center postings; a title can hold more than one, so rows overlap

  1. Technician1,881 · 28.0%
  2. Manager1,701 · 25.3%
  3. Engineer1,394 · 20.7%
  4. Construction664 · 9.9%
  5. Operations478 · 7.1%
  6. Mechanical386 · 5.7%
  7. Security146 · 2.2%

What it shows

Running the buildings (technicians, managers, engineers, operations) is most of it; construction titles number 664. Most of the build-out hires under other titles: see the next figure.

Method and limits

One count per word, on the title only. "Data Center Technician, Data Center Operations" is in both the technician and the operations rows, so the rows add up to more than the total.

Source: Metix AI Platform, jobs, 2026-09-30.

Querydata-center-titles.json
POST /v1/jobs/query · queries/data-center-titles.json
{  "where": {    "all": [      {        "field": "title",        "match": "data center"      },      {        "field": "location.country",        "eq": "United States"      }    ]  },  "size": 1}

More than one in ten US construction lead postings names data centers

The share of each group of postings whose description names data centers

  1. Construction leads, US2,800 / 26,211 · 10.7%
  2. Electricians, US152 / 5,796 · 2.6%
  3. Electricians, Arizona18 / 125 · 14.4%
  4. Electricians, Texas22 / 627 · 3.5%
  5. Electricians, Virginia8 / 349 · 2.3%
  6. Electricians, Ohio5 / 236 · 2.1%
  7. Electricians, California6 / 285 · 2.1%
  8. Electricians, Georgia2 / 280 · 0.7%

What it shows

Of construction manager, superintendent and project engineer postings, 2,800 (10.7%) name data centers, 504 of them in Texas and 161 in Virginia. In a read of 50, 45 name data centers at least twice: the job is the project, not a line in a list of sectors.

Among electricians it is 2.6%, but 18 of 125 in Arizona, 14.4%.

Method and limits

All three spellings count: data center, datacenter and data centre. In a read of 60 electrician postings, 55 do name data centers.

Source: Metix AI Platform, jobs, 2026-09-30.

Queriesconstruction-leads.json · electricians.json · mentions-data-centers.json
POST /v1/jobs/query · queries/construction-leads.json
{  "where": {    "all": [      {        "any": [          {            "field": "title",            "match": "construction manager"          },          {            "field": "title",            "match": "superintendent"          },          {            "field": "title",            "match": "project engineer"          }        ]      },      {        "field": "location.country",        "eq": "United States"      }    ]  },  "size": 1}
POST /v1/jobs/query · queries/electricians.json
{  "where": {    "all": [      {        "field": "title",        "match": "electrician"      },      {        "field": "location.country",        "eq": "United States"      }    ]  },  "size": 1}
POST /v1/jobs/query · queries/mentions-data-centers.json
{  "_note": "Added to another population: postings whose description names data centers, in any of the three spellings.",  "where": {    "any": [      {        "field": "description",        "match": "data center"      },      {        "field": "description",        "match": "datacenter"      },      {        "field": "description",        "match": "data centre"      }    ]  }}

Run it

Three ways in. Each says what it costs before you start.

1 API Credit buys 25 search results or 5 full records; $1 buys 30 API Credits.

Run it in your agent

80 to 95 API Credits$2.67 to $3.17Fits in the 100 free API Credits

Your agent stops and asks before spending more than 110 API Credits.

Your agent follows the prompt step by step: it reads the rules, runs the counts, checks the definitions the prompt asks it to check, and writes the files and charts. Use an agent that can write files, such as Claude Code or Codex.

Set up onceKey, connection, and a free check. Skip this if your agent already reaches the Platform.

  1. 1Get a key

    Create a key on the Metix AI Platform →

    New accounts get 100 API Credits once, valid for 30 days. Set the key in the shell you start your agent from, or add the line to ~/.zshrc or ~/.bashrc so every new terminal has it:

    Shell
    export METIX_KEY=metix_xxxxxxxx

  2. 2Connect your agent

    Claude Code

    Registers the Platform for every project. Start claude in any folder and the ten metix tools are there.

    Shell
    : "${METIX_KEY:?run step 1 first}" &&
claude mcp add --scope user --transport http metix \
  https://mira-api.metix.ai/mcp \
  --header "Authorization: Bearer $METIX_KEY"
    MCP setup guide →

    Codex

    Registers the same server. The key stays in your environment instead of the config file.

    Shell
    codex mcp add metix \
  --url https://mira-api.metix.ai/mcp \
  --bearer-token-env-var METIX_KEY
    MCP setup guide →

    Skills

    Four skills that teach any agent the Platform's endpoints and query rules, for agents without MCP. The installer starts with none ticked: press space on each, then enter.

    Shell
    npx skills add MetixAI-Official/metix-skills
    Skills install guide →

    Other MCP

    Point the client at this endpoint over streamable HTTP with both headers; without the Accept header the server answers 406. Older clients use /sse on the same host.

    Endpoint and headers
    https://mira-api.metix.ai/mcp
Authorization: Bearer <your key>
Accept: application/json, text/event-stream
    MCP setup guide →

  3. 3Check the setup

    Ask this first. It reads your balance and the field list, runs no search, and costs nothing:

    Prompt for your agent
    Use the Metix AI Platform to check my key status and read the contract; both are free. Then tell me my API Credit balance and which datasets I can query. Do not run any search.

4Paste the prompt

Start your agent in an empty folder, then paste. It writes its files there.

The question

Answer one question with the Metix AI Platform: where in the US, and by whom, are companies hiring to build and run data centers? Work only through the public Platform (REST at https://mira-api.metix.ai, the MCP server, or the metix-skills) with the key in METIX_KEY, and never print the key.

  1. 01Read before querying

    Call GET /contract (free) and use only querySpecByEntity.job fields. A count with size 1 costs 1 API Credit; reading records costs 1 API Credit per 5. Check the balance with GET /auth/key/status (free) at the start and at the end, and stop and ask before the run passes 110 API Credits.

  2. 02Population

    Open US postings whose title matches "data center". Read 25 titles to confirm they are data center jobs.

  3. 03Count by state

    One count per state for the 30 largest, then one count for every other state (not in the list). The national total is their sum; do not rely on a single total, which can be banded.

  4. 04Count by employer

    Match company.name exactly as the records spell it; read a few records to learn the spellings (Amazon posts under two names). Report each employer's share of the national total, and Amazon's count in the two largest states.

  5. 05Count by role

    One count per word in the title: technician, manager, engineer, construction, operations, mechanical, security. A title can hold more than one word; say that the rows overlap.

  6. 06The trades

    Count US electricians (title "electrician") and construction leads (title "construction manager", "superintendent" or "project engineer"), then the same with a description that names data centers ("data center", "datacenter" or "data centre"). Read 30 of each matched set to check that the mention is about data center work, not a list of sectors, and report how many were.

  7. 07Clean

    Reposts and multi-city listings are not collapsed; say that the numbers measure hiring activity, not seats.

  8. 08Outputs

    Write data/states.json, employers.json, roles.json and trades.json with "unit": "jobs", the snapshot date and each row's count and share. Report the balance before and after as the cost.

  9. 09Charts

    States as bars scaled to the national total, the largest highlighted and every other state last. Employers and roles as shares of the national total. The trades as the share of each population that names data centers.

  10. 10Limits

    Titles only for the main population. One day, not a trend: open postings skew recent, so posting dates say nothing about growth.

Read PROMPT.md →

What you get

The aggregate files and the chart, a note on what the audits found and what they changed, and the API Credits the run spent, read from the balance before and after.

A call that returns 402 insufficient_quota means the key works and the balance is empty.

Reproduce the numbers

68 API Credits$2.27Fits in the 100 free API Credits

A short standard-library Python script sends the committed queries as counts and writes the aggregate files this page is built from. It needs Python and METIX_KEY set in the shell (step 1 of the agent path); your agent can run these lines for you as well.

Shell
curl -fsSL https://platform.metix.ai/casebook/source/data-center-jobs-us-2026.tar.gz | tar xz
cd data-center-jobs-us-2026
: "${METIX_KEY:?set METIX_KEY first}" && python3 cases/data-center-jobs-us-2026/fetch.py

What you get

data/*.json and data/receipt.json. Run git diff cases/data-center-jobs-us-2026/data to see what moved: the numbers should match, apart from what changed in the data since the snapshot.

Adapt it

Costs what your version reads. Write your own ceiling into step 1 of the prompt.

The prompt is the case. Change the parts in this table and your agent answers your question instead, with the same checks and the same way of reporting cost.

To change Edit For example
The industry Step 2 "semiconductor", "battery", "wind"
The trades Step 6 "HVAC", "pipefitter", "lineman"
The breakdown Step 3 Cities within one state, or metro areas

What to ask before running it

  1. Which title, and does it mean anything else?
  2. States, cities or employers?
Read PROMPT.md →

Method and limits

How the population was defined, counted, and checked, and what the numbers cannot show.

Population

Open US postings on the Metix AI Platform on September 30, 2026. The main population has "data center" in the title; a read of 100 titles found every one to be a data center job.

Counting

Counts only, one API Credit each, 68 in the replay, and no posting is read. The national total is 30 state counts plus one count for every other state, so it never depends on a single total that may be banded. The conditions are in queries/ and the state, employer and role lists in fetch.py.

Checking the matches

The agent read 60 electrician postings (55 name data centers in the description) and 50 construction lead postings (45 name them at least twice), to check that a match is data center work and not a line in a list of sectors.

Limits

Reposts and multi-city listings are not collapsed. Employers are matched on exact names, so their counts are floors. One day, not a trend: open postings skew recent, so posting dates say nothing about growth.

The last reproduction

What reproducing this case cost the last time the script ran, read from the Platform's own balance before and after.

Ran on
2026-09-30
Calls
68
Search results
68
Records read
0
API Credits
68

Search results are IDs returned by searches, one per count query and one per match on a full search. Records are postings or profiles read in full: this case reads none.

Making this case cost about 212 API Credits more: the agent's audits, trial queries, and earlier runs that the published replay replaced. You do not pay that again.

Case files: prompt, queries, dataAPI Credits and pricing