FPT SoftwareFHNCertificationsReference architectures
Talk to NGT
AI & Agents · 06 of 6

Forward-deployed engineering

Engineers inside your team until the first agent runs in production, then a clean hand-over.

awsazuregoogle-cloudkubernetesgithubgitlab
Built in your environment, on your stack
01

How an AI engagement runs

Five stages, from strategy to production, with a go / no-go decision after the first working version.

  1. 01Assessment
    • Current state and data
    • Maturity scored on a checklist
    • Target level agreed
    agreed target
  2. 02MVP / PoC
    • A working demo on a real workflow
    • Fail fast, then re-estimate
    go / no-go
  3. 03Planning
    • Master plan and work breakdown
    • Cost estimate
  4. 04Implementation
    • Build to the target level
    • Harness, evaluation and runbooks included
  5. 05Operation
    • Monitor and maintain in production
    • Improve release by release
02

What forward-deployed engineers do

Joined NGT's consulting services in 2026. Engineers work inside your team until the first agent runs in production.

03

Who does what in a forward-deployed engagement

Stage
Your team
NGT engineers
AI tools and agents
01Assess
Your team
  • Name the workflow and its owner
  • Share data and access rules
NGT engineers
  • Map the workflow and data
  • Score maturity, propose a target
AI tools and agents
    02PoC
    Your team
    • Try it with real users
    NGT engineers
    • Build the first agent and its test set
    AI tools and agents
    • Coding assistants draft code and tests
    03Build
    Your team
    • Approve the design and the sign-off steps
    NGT engineers
    • Add guardrails, evaluation, observability, hand-over
    AI tools and agents
    • Agents handle routine cases
    04Hand over
    Your team
    • Run it, or keep NGT running it
    NGT engineers
    • Runbooks, test sets and training
    AI tools and agents

      Questions this raises

      The questions this part raises, and their answers.

      With an assessment and a short PoC on one workflow, with a success measure agreed up front.

      SourcesNGT Service Offering v3.2 (Jul 2026)AI-First DevOps Transformation: Services & Solutions, NGT (Oct 2026)Full SDLC with GenAI & Multi-Agent AI, NGT (Jun 2025)Platform Engineering v1.2, NGT (Sep 2026)FHN AI Service Introduction v0.9.3 (Mar 2026)Executive briefing: Look Forward 2027, FHN (Oct 2026)FPT Smart Cloud: AI Agents, Token Factory and AI Factory introductions (2025–2026)