Engineering the vision. Executing the growth.

A visionary AI leader who pushes boundaries and breaks the norm. Hands-on in agentic AI development, with architectural and recursive systems thinking that bridges business, AI, data and technology.

Enterprise impact

High-profile transformations across telecommunications, finance and infrastructure, delivering measurable business outcomes at scale.

Journey

15+ years in consulting and engineering (Accenture, EY, Deloitte).

  1. 2013

    Foundations

    Deloitte

    Built analytical muscle in regulated industries. Led UK railway and transport analytics for public sector transformation.

    EdgeLearned to navigate complex stakeholder dynamics and translate data into executive decisions.

  2. 2015

    Product & Automation Pioneer

    EY

    Built a global Insurance Premium Tax tracker toolkit and managed EY's Global Tax Risk Analyzer platform. Developed RPA solutions that cut 90% of manual process costs.

    EdgeFirst experience owning commercial outcomes and building at enterprise scale.

  3. 2017

    Enterprise Delivery Lead

    Accenture

    Delivered a £100M+ financial transformation programme. Built a GenAI application that automates data lineage and mapping across 150+ systems for a banking client.

    EdgeMastered enterprise procurement, stakeholder dynamics, and regulated industry delivery.

    BT case study
  4. 2022

    Data & AI Practice Lead

    Accenture UKI

    Built and led 100+ consultant community from zero. Established KPIs, training programmes, and operational excellence metrics.

    EdgeProved I can build teams and processes where none exist.

  5. 2024

    FDE Team Founder

    Accenture

    Founded the Forward Deployed Engineering team. Built partnerships with 20+ AI startups and frontier AI labs including Anthropic. Created engagement frameworks and reusable delivery patterns.

    EdgeAlready know how to embed with enterprises and deploy production AI agents.

    View conference app

Publication

Ideas and perspectives on AI transformation, engineering excellence, and the future of work.

Rx Weekly Digest, latest issue

The Next Bottleneck Is Not Code

Issue 08, 8 Nov 2026, 18 min read

In the first week of November, HumanLayer joined Factory, and Cursor let anyone in an organisation build and publish live apps. Weeks earlier, at OpenAI’s DevDay, its models had helped tune the software that runs them. Follow one digital team’s Friday through three stages of what AI SDLC may mean for you.

Agents, Engineering, Operating Model, Architecture, Security, Enterprise

Read issue 08

Live from digest.reinventionx.app.

Still from the AI Engineer talk
Talk

Most Enterprise Agentic Projects Are Doomed, Here's Why

AI Engineer, with Jess Grogan-Avignon.

Speaker

Ministry of Justice AI & Digital Professions Conference 2026

AI in public sector transformation.

How this site works

One typed profile feeds this page, llms.txt, profile.json and the structured data that search engines and AI assistants read. A public MCP server answers AI agents, an owner MCP server and a private agent skill keep the facts current, and a job agent scans every morning.

Fig 1Anatomy of jackwang.org
Anatomy of jackwang.orgA split-flap departures unit with its service bay open. One motor and drive shaft turn four word-flap modules: the page, llms.txt, profile.json and the JSON-LD. The bay holds the facts ledger on a tape reel, three guardrail breakers, the binder of the private skill, the guarded approval switch and the job agent's cam timer. The end panel carries the public MCP port, the keyed owner MCP port and the recruiter-agent port with its card reader.PEOPLEPAGE MODELSLLMS AGENTSJSON SEARCHJSONLD
Select a part to hold it9 parts live, 0 next
  1. Live. One source for the page, llms.txt, profile.json and JSON-LD.

  2. Live. Read-only tools any AI agent can call for my public profile.

  3. Live. Owner only. My agents manage facts, CVs and the job search.

  4. Live. The playbook my coding agents follow to maintain my site headlessly.

  5. Live. One statement per fact. Each stays a candidate until I confirm it.

  6. Live. Forbidden-claim, client-name and number checks on every fact.

  7. Live. Agents help me with the process. I decide when to engage.

  8. Live. Recruiter agents read my public facts. I decide on every contact.

  9. Live. Scans public job boards each morning and records matches for me.

Jack Wang

Job search

An agent scans job boards every morning for roles that match my search presets and scores each match for fit. I shortlist the best, the agent drafts each application, and I review and send it myself. The board shows the latest moves.

29 hours Live

  1. 10,935Scanned
  2. 168Matched
  3. 82Scored
  4. 14Shortlisted
  5. 7Prepared
  6. 0Applied

Where the hours come from

  1. Attention12.2 h
  2. Reading8.2 h
  3. Editing8.8 h
  4. Personal productivity0.0 h

How the hours are counted

Attention.
4 seconds per posting title scanned, times scanned (10,935).
Reading.
6 minutes per job description read and scored, times scored (82).
Editing.
75 minutes per prepared application, a tailored CV at 45 minutes and a cover letter at 30, times prepared (7).
Personal productivity.
15 minutes per application form and tracker update, times applied (0).

The baseline is doing by hand exactly what the agent does. Funnel and hours count every posting across 24 scan runs since 25 Oct. Preset match counts are the distinct postings the scans have matched to each preset.

What it looks for

Lead a forward deployed or applied AI team that takes frontier models into regulated enterprises and owns the customer outcome end to end.

Looks for
  • forward deployed
  • fde
  • applied ai
  • solutions architect
  • solutions engineer
  • deployment
  • customer engineer
  • ai engineer
  • field cto
  • head of solutions
Seniority
  • manager
  • head
  • lead
  • director
  • principal
  • staff
  • field
Skips titles with
"intern", interns, graduate, junior, recruit, talent, data center, datacenter, community, sales development

On. 50 title matches on 40 boards, across every scan run.

Jack Wang

Contact

Messages come straight to me, and I review every CV request myself.

Portrait of Jack Wang

London, UK.

Your name, email and message are deleted 90 days after you send it.

Jack Wang

v2, Departures

The Departures redesign: a one-screen title page with a split-flap job board, four deck cards for impact, journey, public work and how the site works, a contact form and a gated CV request, and llms.txt, profile.json and structured data for search engines and AI assistants.

Released 23 Oct 2026, 22:54 UK time. The version you are looking at.

Updates, newest first

  1. The home page with the job board.

    v2.37

    Owner pages load faster.

  2. The home page with the job board.

    v2.36

    Job search screening rules are set from the owner console.

  3. The home page, with Archived in the job board's legend.

    v2.35

    Job search roles that go quiet are archived.

  4. The site footer reading v2.34.

    v2.34

    The job search reads more places for new roles.

  5. The job search sheet's four search tracks.

    v2.33

    The job search page shows its live search tracks.

  6. The site footer reading v2.32.

    v2.32

    The job search follows four refreshed tracks.

  7. The site footer reading v2.31.

    v2.31

    The job search widens its reach, with clearer priorities for the roles it finds.

  8. The job search board on the home page, with a legend for triaged, scored, shortlisted, prepared and applied.

    v2.30

    The job search figures now follow the live record, with a mark for each stage and roles that read whole.

  9. The site footer reading v2.29.

    v2.29

    Job search pipeline improvements.

  10. The departures board on the title page, showing masked live roles.

    v2.28

    The departures board follows the live job search.

  11. The site footer reading v2.27.

    v2.27

    Mobile reading improvements.

  12. The site footer reading v2.26.

    v2.26

    Mobile reading fixes and a lighter page.

  13. The site footer reading v2.25.

    v2.25

    Updates for AI agents and copy edits.

  14. The site footer reading v2.24.

    v2.24

    Copy edits, owner console improvements and back-end AI updates.

  15. The site footer reading v2.18.

    v2.18

    Mobile layout, job board and privacy updates, and copy edits.

  16. jackwang.org v2.10: the enterprise impact card, unchanged on the page while the agent-facing files gained sectors and key strengths

    v2.10

    Release history, profile updates and owner console improvements.

  17. jackwang.org v2.05: the anatomy figure on the How this site works card

    v2.05

    Profile and design updates, and a figure that explains how the site works.

  18. jackwang.org v2.00: the Departures title page with its split-flap job board

    v2.00

    Departures: a split-flap redesign in a Deck layout, with contact and CV requests, and pages written for search and AI assistants.

Jack Wang

v1, v0 launch

A single-page profile generated with v0: hero, enterprise case studies, thought leadership, and three customised pitch pages with their own llms.txt files.

Released 11 Mar 2026, 07:26 UK time. Superseded by a later version.

Copy that has since been corrected is blurred, and the full-page captures are withheld because they named the employers the old pitch pages were written for.

The first screen of jackwang.org v1, with copy that has since been corrected blurred