Blog

Product Rhythms in the Age of AI Part04 | If I Were Running a Product Team Today, I Wouldn’t Run Scrum the Same Way

Product Rhythms in the Age of AI Part03 | I’m Less Interested in Whether We “Use Scrum” and More in Which Rituals Still Earn Their Keep

Product Rhythms in the Age of AI Part02 | AI Won’t Replace PMs First. It Makes Some Work Cheaper Before It Does Anything Else

Product Rhythms in the Age of AI Part01 | It’s Not Agile That’s Aging. It’s Some of the Ways We Deliver

Business Apr 15, 2026

Help Us Understand: What Matters Most in an AI Companion for Elders

We are researching how voice AI companions could meaningfully support elders and their families. Your input — whether you are an elder or a caregiver — will directly shape what we build.

Business Apr 13, 2026

Behind the Trend 01 | How Singapore's Policies Are Turning into Business Opportunities, from Smart Nation 2.0 to NAIS 2.0

Singapore is not only writing tech policy. It is designing the infrastructure for the next round of business formation. This is what I found after reading these documents.

AI Apr 12, 2026

Local LLM Fine-Tuning, Explained Appendix B | Commands, warnings and troubleshooting quick reference

AI Apr 11, 2026

Local LLM Fine-Tuning, Explained Appendix A | Glossary and layer map

when you see a term and want to know what it is, which layer it belongs to, and where you would actually change it, this is where you should be able to find a usable answer quickly.

AI Apr 10, 2026

Local LLM Fine-Tuning, Explained part 09 | How to choose a mainline version: Lexi, adapter variants and the judgement that remains

what should I actually keep?

AI Apr 10, 2026

Local LLM Fine-Tuning, Explained part 08 | From adapters back to Ollama: merge, quantisation and deployment

a finished training run is not the same thing as a finished model

AI Apr 10, 2026

Local LLM Fine-Tuning, Explained part 07 | Training cost on MPS: why things become slow, stuck and out of memory

training is not slow inference. Training is forward pass, loss, backward pass, gradients, optimiser state and parameter updates all competing inside the same space.

What Reality Corrected 05 | Interest is not movement

Near the end of a first-customer-success workshop at Draper U, the speaker drew a ladder on the whiteboard. At the bo…

What Reality Corrected 06 | A better product is not enough if switching still hurts

In another Draper U session, the speaker asked a nastier question than “why would a customer buy?” **If the customer …

What Reality Corrected 07 | Early startup value is not your title. It is how quickly you turn ambiguity into useful work

At Draper U, when Cristina Cordova started talking about her path through Stripe, Notion, First Round and now Linear…

What Reality Corrected 08 | Culture is not what you write on the wall. It is what your decisions repeatedly train people to do

Before Jevan Soo Lenox’s fireside at Draper U started, I had a fairly defensive posture towards the topic. Anything l…

AI Apr 9, 2026

Local LLM Fine-Tuning, Explained part 06 | How to read the training scripts: from `train_lora.py` to `train_partial_ft.py`

an experiment design document that binds together the model, the data, the tokenizer, the training method, the trainable scope and the training rhythm

AI Apr 9, 2026

Local LLM Fine-Tuning, Explained part 05 | What DPO is actually changing

what exactly should the model learn when there is no single canonical answer, only a preference that one answer is better than another?

AI Apr 9, 2026

Local LLM Fine-Tuning, Explained part 04 | What SFT, LoRA and full fine-tuning each actually change

SFT answers how you teach. LoRA answers how you change parameters. Full fine-tuning answers how deep you are willing to cut.

Startup Apr 9, 2026

What Reality Corrected 03 | A pitch deck is not your company story. It is the thing that decides whether an investor keeps reading

At Draper U, Melanie opened her pitch workshop with a line that was more useful than any deck template. Imagine your …

Startup Apr 9, 2026

What Reality Corrected 04 | The market is not too small. You may simply not have earned the right segment yet

In one of the go-to-market sessions at Draper U, the speaker wrote three short lines on the board: - Who you sell

AI Apr 8, 2026

Local LLM Fine-Tuning, from Modelfile and LoRA to DPO part 03 | How the toolchain splits: Hugging Face, Transformers, PEFT, TRL, and Ollama

AI Apr 8, 2026

Local LLM Fine-Tuning, from Modelfile and LoRA to DPO part 02 | How model memory actually splits: context, external memory, parameter memory, and continual learning

AI Apr 8, 2026

Local LLM Fine-Tuning, from Modelfile and LoRA to DPO part 01 | The layer map of a local LLM: from prompts and Modelfiles to model weights

Startup Apr 8, 2026

What Reality Corrected 01 | Before anyone replies, they need to know where to place you

The thing I kept from Rachel Konrad's two Draper U sessions was not a writing trick. It was something more awkward th…

Startup Apr 8, 2026

What Reality Corrected 02 | It wasn't that you were bad at networking. Most networking events are just badly designed

The first time this really clicked for me was not in class. It was after I hosted a founder × investor mixer in the B…

LLM Application Engineering and Governance for PMs Part 4 – Agents Are Not Magic: When to Use Multi-Step Workflows, and When to Stop at Guardrails, KPIs, and Governance

Once tool calling and RAG begin to work, nearly every team develops the same instinct.

LLM Application Engineering and Governance for PMs Part 3 – LLMs Do Not Do the Work Alone: How Tool Calling and RAG Connect Models Back to the Real World

The first instinct many teams have when they start building with LLMs is fairly predictable: write a better prompt, add a few examples, and see whether the model can somehow hold the whole thing toget

After the Pause | On Starting, Pivoting, and Pausing a Venture Over Two Years — Part 07: Going from founder back to candidate was harder than I expected

I assumed one thing would at least become simpler once I came back to the job market after a stretch of building a company.

LLM Application Engineering and Governance for PMs Part 2 － From Prompt to Output Contracts: Why JSON Schemas, Validators, and Retries Are the First Ticket to Enterprise LLM Adoption

- You have already written prompts and started to notice that “looks right to me” and “the system can rely on it” are two very different standards - You are building classification, extraction, summar

After the Pause | On Starting, Pivoting, and Pausing a Venture Over Two Years — Part 06: I Did Not Give Up. I Just Needed to Stay Afloat

In the end, I did not pause the venture because I stopped believing in the problem.

LLM Application Engineering and Governance for PMs Part 1 － Beyond Prompt Engineering: What PMs Are Actually Doing When They Put LLMs into Products and Workflows

- You already use ChatGPT, but you’re now trying to wire LLMs into a product, an internal workflow, or some form of automation - You keep hearing prompt engineering, RAG, tool calling, agents, and it

After the Pause | On Starting, Pivoting, and Pausing a Venture Over Two Years — Part 05: Everyone Knew the Problem. Why Wouldn’t Hoteliers Move?

The response I heard most often during BD was not confusion.

Category Design for Founders Part 2 － How Founders Actually Do Category Design: Problem Framing, POV, Languaging, Fundraising, and Go-to-Market

Part 1 dealt with one thing first: category design is not a naming game, and it is not a grand branding wager every startup should force. At its core, it is about correcting a market that is using the

After the Pause | On Starting, Pivoting, and Pausing a Venture Over Two Years — Part 04: I Thought Product Would Be the Hard Part

At the beginning of my second venture, I placed most of my attention where many founders do.

Category Design for Founders Part 1 － What Category Design Is, and Why Founders Should Not Bet on a ‘Better Product’ Alone

When founders first hear “category design”, the reaction is often immediate.

After the Pause | On Starting, Pivoting, and Pausing a Venture Over Two Years — Part 03: It Was Not a Lack of Direction. Reality Kept Changing the Question

There was a stretch when my pitch deck changed every few weeks.

PM User Research and Fieldwork 04 － Five Participants Is Not a Recruitment Strategy: How PMs Find the Right Users, Screen Them Properly, and Avoid the Wrong Ones

A great many research rounds do not fail at the discussion guide. They fail earlier.

After the Pause | On Starting, Pivoting, and Pausing a Venture Over Two Years — Part 02: The First Thing I Wanted to Build Was Too Big, and Far Too Heavy

The first thing I genuinely wanted to turn into a company was not a small tool.

PM User Research and Fieldwork 09 － A Complaint Is Not Yet a Job: How PMs Confirm JTBD, Switching Moments, and Pain Intensity Without Jumping to Solutions

Professional methods piece / Problem-definition guide

After the Pause | On Starting, Pivoting, and Pausing a Venture Over Two Years — Part 01: I Did Not Suddenly Decide to Start a Company

One of the most common reactions I get, once people learn I went off to build a start-up, goes something like this.

PM User Research and Fieldwork 08 － From Transcript to Insight: How PMs Use Thematic Analysis and Affinity Mapping Instead of Cherry-Picking Quotes

Analysis methods piece / Research synthesis guide

PM User Research and Fieldwork 07 － Interviews and Fieldwork Are Not Casual Conversations: How PMs Facilitate, Observe, Take Notes, and Avoid Distorting the Session

Practical methods piece / Session facilitation guide

PM User Research and Fieldwork 06 － An Interview Guide Is Not a List of Questions: How PMs Design Non-Leading Prompts That Surface Real Evidence

Method piece / Interview craft guide

PM User Research and Fieldwork 05 － Outreach, Screeners, Incentives, and Consent: Turn Research Recruitment into a Repeatable System

Practical method piece / Research ops primer

PM User Research and Fieldwork 03 － Not Every Research Session Is an Interview: User Interviews, Usability Tests, Field Studies, and Diary Studies Serve Different Jobs

Method piece / method-selection piece

PM User Research and Fieldwork 02 － Qualitative, Quantitative, and Mixed Methods: PMs Do Not Need a Side, They Need the Right Question

Method article / concept calibration

PM User Research and Fieldwork 01 － When the Dashboard Speaks but the User Does Not: Why PMs Need User Research to Close Product Blind Spots

Concept calibration / series opener

PM Growth Levers and Monetisation 06 － Growth OS: Mature PMs Do Not Just Generate Ideas—They Turn Metrics, Experiments, Decisions, and Scale into a Repeatable Operating System

A surprising number of teams are not short of ideas. What they are short of is rhythm.

PM Growth Levers and Monetisation 05 － Acquisition Is Not Just Conversion: PMs Should Understand SEO, Paid, Partnerships, Attribution, and Incrementality

When PMs talk about acquisition, the conversation often collapses far too quickly into a few familiar questions: what is CAC, what is ROAS, which channel converts better?

PM Growth Levers and Monetisation 04 － Monetisation Is More Than Slapping on a Paywall: How to Choose Between Fake Doors, Trials, Bundling, and Usage Limits

When teams start talking about monetisation, the conversation often rushes straight to the paywall.

PM Growth Levers and Monetisation 03 － Lifecycle Is Not About Sending More Nudges: Retention Is About Designing a Reason to Return

When teams say they want to improve lifecycle, what they often mean is “we should probably send more messages”.

PM Growth Levers and Monetisation 02 － Value Proposition, Message Tests, and Landing Pages: Many Growth Experiments Fail Because the Promise Is Vague

A surprising number of growth problems are really translation problems.

PM Growth Levers and Monetisation 01 － Activation Is Not Onboarding: PMs Should Start with the Aha Moment, TTFV, and the First Verifiable Value Moment

A lot of teams say they want to improve activation, but what they actually end up working on is onboarding.

PM Product Data and Experimentation 08 － Data Rarely Breaks at the SQL Layer: PMs More Often Fight Identity Issues, Late Events, Bot Traffic, and Failed Rollouts

When PMs see strange numbers, the first instinct is often something like this:

PM Product Data and Experimentation 07 － A/B Testing Is Not Just Hypotheses and P-Values: PMs Need Exposure Rules, SRM, Guardrails, and Validity

A surprising number of experiments do not fail because the idea was weak. They fail for a far more awkward reason.

PM Product Data and Experimentation 06 － Retention, Cohorts, and Segmentation: Knowing Who Stays Matters More Than Watching the Average

A product can look healthier than it really is simply because averages are excellent at cosmetics.

PM Product Data and Experimentation 05 － Window Functions, Sessionisation, and Funnel SQL: Where SQL Starts Becoming a PM Superpower

A lot of PMs learn SQL by learning how to count.

PM Product Data and Experimentation 04 － SQL for PMs: Learn to Pull Evidence Before You Claim to Be Data-Driven

When PMs say they want to become more data-driven, the first instinct is often to ask for more dashboards.

PM Product Data and Experimentation 03 － A Tracking Plan Is Not an Event Checklist: It Is a Data Contract Between PM, Engineering, and Analytics

When teams talk about a tracking plan, they often sound as if they are managing a to-do list.

PM Product Data and Experimentation 02 － Metric Trees Are Not Enough: Why PMs Should Understand Metric Dictionaries, Governance, and the Semantic Layer

When PMs start building real data fluency, one of the first “advanced” tools they tend to learn is the metric tree.

PM Product Data and Experimentation 01 － North Star Is Not Enough: How PMs Use HEART, Guardrails, and Counter-Metrics to Define Success

When PMs first get introduced to product metrics, they usually hear a powerful idea very early on: find your North Star Metric, and the whole team will start pulling in the same direction.

Running a Local LLM on a Mac mini M4 with 16GB RAM: Why I Started with Ollama and Llama 3.1 8B

If you have a **Mac mini M4 with 16GB of unified memory and a 256GB SSD**, the first problem with local models is rarely installation. It is choosing a sensible starting point.

RAG Engineering in Practice 07 – How My Job Agent Turns JDs, CVs, and Rubrics into a Working Evidence Pipeline

If the first six articles in this series were mostly about sharpening concepts, setting boundaries, and clearing technical debris off the road, this one is about putting those ideas back into a real workflow and seeing how they actually hold together.

RAG Engineering in Practice 06 – Taking RAG into Production: Retrieval, Citation, Evaluation, and Observability Are Not Optional

A lot of RAG systems are very good at performing in demo mode.

RAG Engineering in Practice 05 – Qdrant’s JSON 400 Hell: The Problem Is Usually Not Qdrant but the Body You Actually Sent

If you are calling Qdrant from Make or another low-code tool, the thing most likely to drive you up the wall is usually not vector retrieval itself. It is the class of JSON 400s that look utterly unreasonable.

RAG Engineering in Practice 04 – Why Qdrant Throws `Index required`: Payload Indexes, Schema, and Filter Design Explained

The first time you add a filter in Qdrant, there is a decent chance you will run into an error that feels oddly rude:

RAG Engineering in Practice 03 – The Engineering Philosophy of Chunking: You’re Designing Evidence Units, Not Splitting Text

One of the first terms people run into when building RAG is chunking. And that is often where things start to drift.

RAG Engineering in Practice 02 – From Relational to Vector: A Practical Database Selection Map

A lot of database discussions still begin with “SQL versus NoSQL”. That is not wrong, exactly. It is just usually too coarse to be very useful for real engineering decisions in 2026.

RAG Engineering in Practice 01 – RAG Is More Than a Vector Database: Start with the Full System Map

When people first get into RAG, their attention often goes straight to the vector database. That is understandable. It is usually the part that sounds the most new, the most technical, and the most likely to impress someone in a meeting.

OpenClaw Deployment and Configuration Part4

When OpenClaw breaks, how to sort the failure into the correct fault line before you waste hours guessing.

OpenClaw Deployment and Configuration Part3

How to handle permissions, updates, backups, and migration without gradually wrecking a working OpenClaw host.

OpenClaw Deployment and Configuration Part2

After installation, which settings are worth touching first, and which ones should wait.

OpenClaw Deployment and Configuration Part1

Why I usually recommend starting OpenClaw on macOS or a Mac mini before reaching for a VPS or a tiny Linux VM.

OpenClaw Apr 3, 2026

OpenClaw Getting Started 06 | How to Choose Models: OpenAI, Claude, Gemini, Grok, MiniMax, z.ai, and local models

A practical model-selection guide for OpenClaw beginners, covering API providers, subscription-based routes, cost feel, quota risk, agent suitability, and Daniel's real-world setup.

ComfyUI Series 07 | Troubleshooting and common pitfalls: SQLAlchemy, requirements.txt, OOM, red nodes, and why safetensors keeps coming up

If you have made it through the earlier parts of this series, you are probably no longer stuck at “how do I install this”, but rather at questions like these:

ComfyUI Series 06 | How do you install models? Checkpoints, clips, loras, vae, T5XXL, GGUF, and FP8 explained

This is the point where things stop being “I vaguely understand the theory” and become “why is my workflow full of angry red boxes”.

ComfyUI Series 05 | How do you actually choose a model? SD 1.5, SDXL, LCM, Turbo, Pony, Flux, and HiDream explained

Once ComfyUI is up and running, the next thing that tends to melt people's brains is the model list.

ComfyUI Series 04 | Where do you actually get models from? What Civitai and Hugging Face do differently, and how to read a model page properly

Who is this for? Anyone who has ComfyUI running and is now ready to install models, only to discover that one page is full of checkpoints, another is full of LoRAs, and everything has version numbers, strange file names, and far too many download buttons. If you’ve just finished the install, this is the right place to go next.

ComfyUI Series 03 | Don’t just click about blindly: `127.0.0.1:8188`, your first image, tmux, background runs, and what happens after a reboot

Who is this for? Anyone who has managed to install ComfyUI, but is now staring at the interface wondering what on earth to do next. If you haven’t installed it yet, read the previous piece first. This one assumes ComfyUI already launches on your Mac.

ComfyUI Series 02 | Installing ComfyUI on a Mac mini M4: get Python, PyTorch and MPS sorted first, and life gets easier later

Who is this for? Anyone who has decided to use ComfyUI and now wants a clean, workable install on a Mac mini M4 with 16GB. We’re not doing the full model menagerie yet, and we’re definitely not diving into Flux on day one. This piece is about laying a sane foundation.

ComfyUI Series 01 | Picking the right tool first: why I’d start with ComfyUI over A1111, Forge or InvokeAI, and whether a Mac mini M4 with 16GB is enough

Who is this for? Anyone eyeing up local image generation on a Mac mini M4 with 16GB, and trying to work out which UI is actually worth installing first. If you’re new to the Stable Diffusion world, start here, then move on to the installation guide.

OpenClaw Getting Started 05 | What OpenClaw Should Actually Do for You, and What It Should Not

A judgment piece with a practical edge. Built around Daniel’s real path, this article maps OpenClaw’s sweet spot, its costs, its counterexamples, and the jobs that are better handled by Codex CLI or a normal chat UI.

OpenClaw Getting Started 04 | How to Use OpenClaw in Your First Week: The Right Order for CLI, Dashboard, Memory, Browser, and Discord

For readers who have already installed OpenClaw and want a practical first-week rollout instead of a feature dump. Built around Daniel’s real macOS and Codex OAuth path.

OpenClaw Getting Started 02 | Discord, Skills, Memory, Remote Access, and the Safety Baseline

For readers who already have the base install working and now want a usable but controlled OpenClaw setup with Discord, skills, memory, remote access, and safer execution boundaries.

OpenClaw Getting Started 02 | A Recoverable First Install on macOS / Mac mini

A practical macOS install guide for technically comfortable readers who want a boring, recoverable OpenClaw baseline before adding channels, skills, or remote access.

OpenClaw Getting Started 01 | What OpenClaw Actually Is: A Self-Hosted Agent Gateway, Not a Stronger Chat UI

A technical introduction for readers who are comfortable with the terminal and want a correct mental model of OpenClaw before installing it.

AI Agentic Workflow Series 7 — How AI connects to the world: MCP will not replace workflows, it is the middle layer that lets agents use tools safely

MCP is not a workflow killer, and it is not merely a trendier wrapper around webhooks or function calling. Its real role is to expose tools, data, and prompts through a standardised interface so agents can connect to external systems more safely and consistently.

AI Agentic Workflow Series 6 — When should you use a Workflow, and when do you actually need an Agent?

Not every multi-step system is an agent, and a human approval step does not magically make one. This article combines Daniel’s practical heuristics with Anthropic and OpenAI’s guidance to offer a more useful way to decide between workflows and agents.

Building AI Skills Series Part 3: Designing a skill layer from existing tools, using a Make-based job workflow as the example

Building AI Skills Series Part 2: Why ChatGPT should not connect directly to low-level tools

Building AI Skills Series Part 1: What is the actual difference between a Skill, Tool, MCP, Runtime, and Orchestrator?

AI Agentic Workflow Series 5 — Why I Eventually Replaced the Make MCP Server with My Own Gateway Server

v2 solved the “Make-first workflow as execution engine” problem. v3 takes the next step and moves the capability surface the model actually sees from the Make MCP server into a self-hosted FastMCP skill gateway.

MCP Engineering Deep Dive 04: Skills Are Not the Same Thing as Runtime Tool Selection

Having skills in your repo does not mean the runtime will actually route by them. In practice, the model is far more influenced by the tool surface you expose, the live inventory a session can see, and whether your descriptions, schemas, and instructions line up.

MCP Engineering Deep Dive 03: Contract Design, Schema Discipline, and Versioning Decide Whether Your Tools Can Grow Up

The hardest MCP servers to evolve are not the ones with too many tools. They are the ones with loose contracts. Once you have multiple clients, public tools, and helper tools, schema discipline becomes the real stability layer.

MCP Engineering Deep Dive 02: Security, Auth, and Public Server Hardening Are Product Boundaries, Not Patches

If your MCP server is public, can write data, or touches real users, security cannot be the last middleware you add. It has to shape your tool surface, auth flow, authorization scope, and response discipline from the beginning.

MCP Engineering Deep Dive 01: Transport and Remote Deployment Are Part of the Design

Part 01 of the MCP Engineering Deep Dive series.

Build Your Own MCP Server — Part 5: The Pitfalls and Operations Manual for Oracle VM and FastMCP

Part 5 of the MCP Server build series.

Build Your Own MCP Server — Part 4: What Skills Are, and How to Design a Skill Layer

Part 4 of the MCP Server build series.

Build Your Own MCP Server — Part 3: How to Compare MCP Frameworks and Choose One

FastMCP, MCP Framework, xmcp, and Spring AI MCP Server are not four skins over the same idea. The real question is which one fits your language habitat, deployment model, auth pressure, and long-term operating style.

Build Your Own MCP Server — Part 2: How to Deploy a Public MCP Server on Oracle VM

Using Oracle VM, Cloudflare, nginx, and FastMCP to turn an MCP server that merely runs in a terminal into a public `/mcp` endpoint that ChatGPT can actually reach.

Build Your Own MCP Server — Part 1: What MCP Actually Changes

This is not just about making tools callable. It is about redrawing the line between reasoning, capability exposure, and execution.

Refactoring a Make-first workflow into an MCP execution engine: a practical migration from toolisation to contracts

How I migrated a Make-first job agent into a clean MCP execution engine, moving from flow-driven design to contract-driven architecture.

AI Agentic Workflow Series 3 — How to Build a Context-Aware Job Agent in Make — Part 2

Part 2 of the job-agent build log: fast scoring, RAG-based deep analysis, and guardrails that turn shortlisted roles into decision-quality output.

AI Mar 31, 2026

AI Agentic Workflow Series 2 — How to Build a Context-Aware Job Agent in Make — Part 1

Part 1 of a practical build log on creating a context-aware job agent in Make, covering LINE intake, continuation handling, routing, and recent-job scraping.

AI Mar 31, 2026

AI Agentic Workflow Series 1 — How I Choose Between LangGraph, n8n, and Make: Why I'm Backing Make for Now

A practical comparison of LangGraph, n8n, and Make - and why, for my current workflow shape, I'm choosing Make first.

Murmurs Mar 28, 2026

Welcome to danielcanfly

A brief introduction to this blog and what you can expect to find here — the intersection of Web3, AI, business, and travel.