Hassan Syed · Applied AI writing
Ideas tested in the real world.
Short field notes and long-form articles on enterprise AI, architecture, product delivery, governance, and the changing craft of software engineering.
Latest insight
6 Sept 2026
AI engineering
Read the articleThe Eval Score That Went the Wrong Way
We thought an uneven AI score might be ordinary variation. Repeated evals told a different story—and showed us how an AI coding harness can help investigate quality without replacing human judgement.
The archive
Writing for leaders and builders
Read NeoSpark articles here, with links to the original publication when an external source is available.
Field note · 3 Feb 2026
Containing autonomous agents with a Control Tower
A practical hub-and-spoke experiment using isolated agents, policy enforcement, risk assessment, audit trails, and human approval for consequential actions.
Read on LinkedInLong-form article · 1 Feb 2026
The Missing Discipline in AI-Assisted Software Development
Why AI agents should execute a complete, testable design rather than invent the product while coding—and a six-phase pipeline for disciplined delivery.
Read on LinkedInLong-form article · 26 Jan 2026
Agile is Dead. Long Live AI-Driven Waterfall.
A practical account of building a production charity platform in three days with an AI-driven, evidence-led workflow that keeps Agile feedback where it matters.
Read on LinkedInLong-form article · 30 Nov 2025
Solving AI Hallucinations with Context-Rich Discovery Tools
How context-rich discovery, data-quality checks, human verification, and MCP-based tooling can turn unreliable AI answers into grounded, traceable insights.
Read on LinkedInLong-form article · 13 Oct 2025
Cursor AI Said 'It's Working' 15 Times Before It Actually Worked - My Gen AI Development Reality Check
Practical lessons from building and deploying with Cursor and multiple frontier models: verification, testing, context, and human judgement still determine whether software works.
Read on LinkedInLong-form article · 8 Oct 2025
The Reality Check: When Your AI Plans Meet Enterprise Complexity
Why promising AI pilots collide with enterprise data, integration, governance, and delivery realities—and how disciplined AI engineering closes the gap.
Read on LinkedInLong-form article · 30 July 2025
When AI Builds It, Tests It, Deploys It, and Uses It, like a human — I Just Watch - Part 1
A hands-on account of using AI agents across software design, coding, testing, deployment, and persona-based product validation while retaining human oversight.
Read on LinkedInLong-form article · 16 May 2025
Do AI Coding Assistants Perform Differently With Different Tech Stack?
How language, framework, architecture, documentation, and test quality can affect the performance of AI coding assistants across different technology stacks.
Read on LinkedInLong-form article · 9 May 2025
AI Dev with Cline/Cursor - Efficient Context Management Is The Key
Why large codebases challenge AI coding assistants—and how feature-based modularisation, bounded contexts, and focused context improve accuracy, speed, and cost.
Read on LinkedInLong-form article · 18 Mar 2025
Essential Eight Compliance Assistant - A Free GPT that Makes Your Solution Architecture Assessment Easy!
A free custom GPT that helps architects and engineering teams assess solutions against the ACSC Essential Eight maturity model and identify practical compliance gaps.
Read on LinkedInLong-form article · 6 Mar 2025
AI Showdown: ChatGPT vs. Grok-3 vs. DeepSeek vs. Claude – Who Makes the Best Charts?
A practical comparison of ChatGPT, Grok 3, DeepSeek, and Claude using the same data-visualisation task—and what happens when the models collaborate.
Read on LinkedInLong-form article · 22 Feb 2025
Comparing AI-Generated Spaceflight Animations: Grok 3 vs. DeepSeek, ChatGPT, and Claude AI
A comparison of Grok 3, DeepSeek, ChatGPT, and Claude generating a 3D Mars mission animation, assessed for execution, visual quality, and scientific accuracy.
Read on LinkedInLong-form article · 21 July 2024
Revolutionising Customer Support with AI & NLP
How AI, natural-language processing, and retrieval can reduce manual searching and help customer-support teams deliver faster, more consistent answers.
Read on LinkedInLong-form article · 2 June 2024
RAG (Retrieval-Augmented Generation) For Dummies. Demystifying A Key Design Pattern For Developing Enterprise AI Applications
A plain-language guide to Retrieval-Augmented Generation, using a tender-discovery example to explain retrieval, augmentation, generation, grounding, and enterprise value.
Read on LinkedInLong-form article · 28 May 2024
Chain-of-Thought, Tree-of-Thought, and Graph-of-Thought Prompting for Health Triage at Home - A custom GPT App
An exploration of Chain of Thought, Tree of Thought, and Graph of Thought prompting for a home health-triage assistant, including benefits and safety limits.
Read on LinkedInLong-form article · 25 May 2024
Birbal - Your Trusted Advisor GPT - Powered by GPT-4o
A custom GPT inspired by Birbal demonstrates how a trusted AI adviser can combine structured personal context, thoughtful questions, and a distinctive voice.
Read on LinkedInLong-form article · 11 May 2024
Great Minds Lounge: Want to tap into Gates, Musk and Jobs' minds? Check out this GPT App
Great Minds Lounge experiments with a panel of AI personas inspired by well-known business and technology leaders to explore ideas from multiple perspectives.
Read on LinkedInLong-form article · 8 May 2024
Elon Musk - Be My Guest : How AI Personas Can Improve Productivity in the Workplace
An experiment with AI personas explores how role-specific context, source material, and transparent boundaries can support workplace thinking and productivity.
Read on LinkedInLong-form article · 5 May 2024
Delta Lake :The Time Traveller's Data Guide
A practical introduction to Delta Lake, showing how transaction logs, schema controls, time travel, and medallion architecture improve modern data platforms.
Read on LinkedInLong-form article · 1 May 2024
Beyond the Hype: Real-World Application of ChatGPT-4 with my BYD Seal Assistant
How a personal BYD Seal assistant turned EV ownership questions into a practical conversational product experiment powered by ChatGPT-4.
Read on LinkedInLong-form article · 10 Sept 2018
FlyView 3.0 Is Here — Using SharePoint Is Much Easier Now
FlyView for SharePoint 3.0 adds support for the modern SharePoint experience and faster navigation across sites, libraries, documents, and site collections.
Read on LinkedInLong-form article · 28 Sept 2015
The SharePoint Governance Puzzle
A practical guide to controlling SharePoint sprawl through ownership, lifecycle management, information architecture, automation, and measurable governance.
Read on LinkedIn