Skip to content
Nikhil Pande

Practical GenAI

Using GenAI to Improve Delivery

I use GenAI as part of my everyday delivery process rather than treating it as a separate technology exercise. It helps me research, structure requirements, prepare documentation, analyse information and identify opportunities to improve repetitive processes. The value comes from combining AI speed with domain context, source validation and human judgement.

01 / 06

Discover

Research and preparation before decisions are made.

  • Research unfamiliar subjects
  • Compare possible approaches
  • Review documentation
  • Identify the important questions
  • Explore technologies and vendors
  • Prepare for technical discussions

Research faster. Structure clearly. Deliver consistently. Improve continuously.

The cycle in detail

  1. Discover

    Research and preparation before decisions are made.

    • Research unfamiliar subjects
    • Compare possible approaches
    • Review documentation
    • Identify the important questions
    • Explore technologies and vendors
    • Prepare for technical discussions

  2. Structure

    Turning unstructured information into working documents.

    • Requirements and technical briefs
    • User journeys and process flows
    • Project plans and acceptance criteria
    • Risk summaries
    • Meeting actions and decision documents

  3. Deliver

    Supporting the day-to-day mechanics of delivery.

    • Project documentation
    • Stakeholder communication
    • Test preparation
    • Release and operational instructions
    • Issue summaries and knowledge sharing

  4. Analyse

    Making sense of large or messy information sets.

    • Review large information sets
    • Identify patterns and trends
    • Summarise operational data
    • Investigate possible issues
    • Compare proposals and spot gaps
    • Improve management reporting

  5. Optimise

    Reducing repetitive work and building reusable processes.

    • Reduce repetitive work
    • Improve workflow efficiency
    • Create reusable processes
    • Faster internal responses
    • Better customer lifecycle communication
    • Identify automation opportunities

  6. Validate

    Keeping judgement, accuracy and accountability human.

    • AI output is reviewed
    • Human judgement owns decisions
    • Facts checked against reliable sources
    • Sensitive information protected
    • Automation must solve a clear problem
    • AI never hides accountability

Hands-on

GenAI Experiments and Workflows

Practical internal workflows and personal systems — described exactly as what they are, not as products. Each one earns its keep by removing real friction from real work.

Jerry — Personal AI Assistant

Personal workflow

Jerry is a developing personal AI assistance workflow designed to help organise information, prepare work, track follow-ups and reduce the time spent switching between routine tasks.

What it does

  • Daily organisation and information management
  • Research support and preparation
  • Reminders and follow-up tracking
  • Reducing context-switching between routine tasks

What it deliberately isn't

  • Not autonomous — Jerry supports my work, it does not act unsupervised
  • A personal workflow, not a commercial product
  • Every meaningful decision stays with me

GenAI Assisted eSIM Management Reporting

Internal workflow

Built with Claude

I use Claude to support the preparation and improvement of management reporting for eSIM operations. The approach helps organise information, identify important trends, create clearer summaries and present operational data in a format that is easier to review and act on.

What it does

  • Structuring operational information into clearer dashboards and summaries
  • Identifying trends worth management attention
  • Decision-focused reporting formats
  • More consistent reporting across periods

What it deliberately isn't

  • Reporting support, not a real-time AI system
  • Figures and conclusions are verified before they are shared

AI Assisted Morning Email Brief

Personal workflow

Built with Google AI Studio and connected email workflows

I have worked on an AI assisted morning briefing approach that turns a large volume of email updates into a structured summary of priorities, important decisions and actions requiring attention.

What it does

  • Reviews relevant emails each morning
  • Surfaces priorities, decisions and follow-ups
  • One structured overview instead of inbox archaeology

What it deliberately isn't

  • The brief supports human review — it does not reply to emails
  • Sensitive emails are never auto-actioned
  • Important decisions remain with me
  • Access and privacy are handled responsibly

Validate

AI speed, human judgement

Every GenAI workflow above runs inside the same guardrails.

  • AI output is always reviewed before it is used.
  • Human judgement remains responsible for decisions.
  • Important information is checked against reliable sources.
  • Sensitive information must be protected.
  • Automation should solve a clear operational problem.
  • AI should not hide accountability.