Founder's Journel

Closing the Gap Between First and Second-Order Thinking in agentic workflows

Tue, 14 Apr 2026 00:00:00 GMT

I was recently speaking with a Principal at KKR, and one part of that conversation has stayed with me.

We were discussing why agent workflows look promising in demos, then break down in real business operations. He articulated the problem very clearly through first-order and second-order thinking.

First-order is what people usually say: "move invoices from one folder to another and process them."

Second-order is the layer they rarely say out loud: do not process duplicates, validate critical values, flag invalid status patterns, and handle messy exceptions before they become risk.

Most operators in front and back office teams are experts in finance, economics, and operations. They are not trained in computer science, prompt design, or AI failure modes. So asking them to define production-grade agent logic from scratch is not realistic.

When teams only provide first-order instructions and expect second-order outcomes, the workflow becomes unreliable. That is usually when the label becomes "AI hallucination," when the real issue is missing operational logic.

This is the kind of design principle that can change how teams solve problems with agents in industry. At Titan, we believe the gap between first-order and second-order thinking should be closed through the product experience.

Titan is a data reasoning and provenance layer for regulated workflows, designed to make operational decisions explicit, traceable, and dependable.

The product question for us is not, "How do we improve the prompt box?" It is, "How do we help operators surface the logic they already hold in their heads through experience?"

Our current approach is to make workflow definition a structured dialogue, not a single instruction.

What cannot fail?
What counts as invalid?
What should happen when data conflicts?
What needs human review before completion?

If we get this right, agent workflows become more authoritative and more useful for everyday teams. Not because everyone became an AI expert, but because the system helped make second-order logic explicit before automation runs.

For Titan, this is a core design direction. Build with the operator's reality, then formalize the hidden guardrails that make the work safe.

Founder Journal: On Gravity

Fri, 13 Feb 2026 00:00:00 GMT

Early startup life can feel like the myth of Sisyphus: pushing a boulder uphill every day, only to return and do it again the next morning. Most founders know that cycle well, and it is usually defined by high effort with very little compounding.

Then, slowly, the question changes. It is no longer just "can I keep pushing?" It becomes "is this starting to pull?" That shift matters because a company does not grow simply because the founder can explain it well. A company grows when people start to feel pulled toward it.

That pull is gravity. In startup terms, gravity is momentum made visible. It is repeated evidence that reality is forming around an idea through consistent shipping, clearer progress, and signals that move the work from interesting to increasingly inevitable.

What Creates Gravity

Four forces create that pull:

Clarity. Focus. Momentum. Ownership.

I did not learn these from a framework. I learned them by watching successful startups create their own gravity, and by getting parts of this wrong myself. I am still working these out in real time.

Clarity

Gravity starts when the story gets simpler, not smaller in ambition, but sharper in problem definition. Optional problems tend to get polite interest, while structural problems attract real commitment.

This came from observing companies that seem to create gravity around them: they land on a clean vision that other people can repeat accurately, so conviction spreads beyond the founding team.

When the pain sounds unavoidable and recurring, the company starts to feel necessary rather than nice to have.

Focus

Breadth can sound impressive at first, but constraint builds trust over time. When a company tries to solve everything, it often signals drift. When it commits to one painful wedge and executes there, it signals seriousness.

I learned this the hard way at DoorFeed and in my own earlier founding journey. At DoorFeed, we tried to replace the whole residential real estate layer to package assets into investable portfolios. It was too many moving pieces and too much operational pain, and it cost the business over two years in the early stage.

Focus does not reduce ambition. It makes ambition believable to the people who matter most.

Momentum

Momentum is not just a headline metric. It is pattern confirmation. You hear the same pain repeatedly across conversations, the product shape moves closer to real workflows, and decisions shift from hypothetical to unavoidable.

Very early on, this often feels like swimming against the current. If you have surfed, you know that paddling out is hard and unavoidable if you want any chance of reading the break properly. In startup terms, that is consistent effort when it feels like nothing is working, so you do not crash out before momentum appears.

At that point, momentum stops being noise and starts looking like direction repeated in public.

Ownership

Exceptional people do not join to observe progress. They join to author it. A company that presents itself as complete leaves little room for ownership, while a company that exposes hard, unresolved work creates a real invitation.

This has been a very clear lesson for me, and I am still navigating it as a solo founder. Inspiring action seems to come from the combination of clarity, focus, and momentum, plus knowing what to say and when.

In practice, people commit where they can become necessary.

The Shift

This is the most important shift to work on, and it is more psychological than practical. You feel it when you stop sounding impressive and start sounding inevitable. That is when gravity takes effect.

Build for the Model's Ceiling, Not Its Cost

Thu, 04 Dec 2025 00:00:00 GMT

I was speaking with the CTO of Granola recently and asked him a simple question. “How do you consistently get the strongest performance out of your agents?” And he replied with an answer that reframed a lot for me.

To give a little bit of context, I've been building agent workflows to be maximally efficient. For example, I would maximise token efficiency at all costs. So trying to get the minimal amount of token usage for generating a response, and some of the time this works, but the result is not consistent.

He said that he prefers to max out the model's context window as much as possible. "It doesn't have to reach to the full amount," he said, "but if it was on a scale, you'd look to medium to high context window." Because you're prioritising giving the user the best experience from the model's capability.

Granola have just raised $60M so that they can give that best user experience to most of the users for low cost or even free. And they're happy to spend that amount to improve the experience, improve the brand, improve their reputation while growth is still happening.

There's also an important point that I came too realise, through implementing the strategy. Not only are you maximising for the current model, but you are also maximising the models' capabilities. This means that when eventually a new model comes out (which will not be too long), you already maximally fit for the current model. The token usage becomes more efficient because over time, models become more efficient. The teams that are building these models are always looking to improve and make them more efficient, the token window becomes less expensive over time.

Actually making these really bold bets in the early stage of this nascent industry and also in the early stages of a company actually provides a really hidden, powerful benefit of:

Delivering an exceptional user experience immediately.
Positioning your product to become dramatically more efficient as model costs drop.

The founders who build like this will dominate the next cycle. They will control more of the market. attract better customers. and develop a stronger brand because they made the right bet early. The founders who optimise for minimal token usage will save a little money now and lose a lot of ground later.

AI capability is accelerating. If you want to lead, build for where the curve is going, not where it is today.

Founder Journal: On Pitching

Tue, 18 Mar 2025 00:00:00 GMT

I have failed to pitch correctly more than anyone can count.

Some got meetings but no second call. Potential customers said they understood the product, then did nothing. I used to think the problem was clarity. If I could only explain the system well enough, people would get it.

That was wrong.

The Room That Taught Me

I remember one pitch where I was walking a room of execs through the architecture. Midway through I noticed them on their phones. One of them leaned forward and said: "I have a busy schedule, can you get to the point?"

I had been clear. I had been precise. And it had not mattered.

The lesson was brutal: clarity alone does not create engagement. There has to be something for the audience to feel.

That moment stayed with me. It was the beginning of understanding why my earlier pitches had fallen flat.

What I Was Doing Wrong

I was starting with the thing I knew best: the product.

"We are an X platform that does Y."

Nothing wrong with that sentence in isolation. Nine times out of ten it communicates what you do. The problem is that it gives the room nothing to feel. No stake. No reason to lean in. I had optimized for understanding and skipped the step that actually makes people care.

Great pitches do not start with what. They start with why.

Not "why we exist" in a mission-statement way. Why this problem matters. Why it is urgent. Why the person in front of you should care before they have heard a single feature.

I learned that the hard way. I am writing it down so I do not forget.

Two Kinds of Emotional Trigger

Over time I started to notice a pattern in the pitches that worked.

When people leaned in, it was because something had landed emotionally. I began to separate those moments into two types.

Direct: You speak to a pain, a belief, or an experience that is personal to them. They feel it in the room. It is about them.

Indirect: You speak to a problem they know exists from the industry, from others, from the news, even if it is not their own story. They are motivated because they recognize the problem and want to be part of solving it.

Neither is better. The work is to know which kind of audience you have, and to aim the opening at that. I had been pitching the same way to everyone. That was another mistake from the earlier attempts.

Knowing Your Audience

The challenge is always the same: who is in the room?

Are they living the problem? Then the direct trigger is available. Are they observing it, managing it, or investing in it? Then the indirect trigger is the one that will land.

This does not come from the deck. It comes from listening before you pitch. From research. From questions that help you place them without giving away the idea. When I had no chance to do that, cold events, networking, I learned to default to why. Why this problem, why now. It is the safest opening when you do not yet know how they relate to the problem.

Why Before What

A "what" statement is: "We are an X platform that does Y." Practical. Clear. Emotionally flat.

A "why" statement answers: why are we solving this in the first place? It connects to something that makes the listener want to hear more.

Example:

"The power demand for AI compute is increasing exponentially and data centres are producing extraordinary amounts of heat. With models getting larger and more compute-intensive, the problem will only get worse. That is why we are building next-generation hardware that is more energy efficient and can handle large-scale compute in one chip."

I could have led with the chip. I could have led with the platform. Nobody would have felt anything. By starting with the problem, the trend, the stakes, the room has a reason to care before the product appears.

The aim is to get them to the edge of their seat, make them ask for the details.

What I Would Tell My Earlier Self

Pitching is more than one lesson. There is timing, pacing, and the psychology of how people decide. But if I could only pass on one thing from the attempts that failed, it would be this:

Do not lead with what you built. Lead with why the problem demands that it exist. Give the audience something to feel before you ask them to think.

It's difficult and you will not get it right on the first try. This is very much an iterative process. In practice, you know it works when you get a follow on conversation. Experiement with your audience and see what works best.

Tenets of 2024

Tue, 31 Dec 2024 00:00:00 GMT

Every year, I like to reflect on the past twelve months and approach it as an observer devoid of bias.

For me, this reflection time is very important, and I treat it like a ritual. It often takes the shape of a solo trip or adventure into nature, honing two tools, a pen and a notebook.

This year, I stumbled across a theme for my reflections, and they decided to take the form of tenets or rules. Some lessons have come from the observations of mentors, friends and colleagues. Others have shown themselves as adverse reactions to behaviour, and ultimately only shown themselves through hindsight.

It is a consistent effort to ensure that we can adopt lessons from others past experiences without living them ourselves.

Admittedly, this is something that I'm not perfect at, but I'm working on it. Here are ten tenets that have stood out to me this year.

Tenets

I can expand my mind beyond any restrictions or rules.
Treat feedback as a gift; whether it is a criticism or a positive, there is something you can learn from it.
Only prove your love, not your worth.
Protect those who can’t, ignore those who won't, and enable those who want.
Stay true to your values, and write your obituary to keep your actions accountable.
You can win regardless of your emotional, physical, and external setbacks. Working hard to overcome these challenges is the difference between winners and losers.
Approach each day like an athlete, show up with a winning mindset, and approach each challenge enthusiastically. Train hard, listen to experience, rest and recover.
Value your friendships, celebrate those around you who are exceptional and work hard.
Recognise the thoughtful actions of others, and be generous with your actions towards them.
The most essential thing in life is your relationships; therefore, be super selective with the people you associate with, especially the person you choose to be your life partner.

If you found these tenets helpful, you can use them as a framework to reflect on your own year. If that is the case I'm happy they inspired you.

Creating Your Own Tenets

If you would like to create your own tenets, the following steps are a good place to start:

Look back on your year with honest eyes, and ask yourself: what principles have guided you? What lessons have shaped you? What values do you want to live by?

If patterns emerge, if themes reveal themselves, write them down. Create your own tenets.

And if you do, I'd love to hear them. Share what you've discovered, what principles have become your compass, or what reflections have shaped your thinking.

There's something powerful in seeing how others navigate their own growth. Your insights might just inspire someone else to start their own reflective journey.

Breaking Free from the Spreadsheet Prison, The Future of Intelligent Data

Tue, 03 Dec 2024 00:00:00 GMT

Working in the investment industry, I saw firsthand how fragmented and siloed data could slow decision-making to a crawl. Multi-million-dollar projects were still tracked in Excel sheets, which were isolated, rigid, and nearly impossible to integrate across teams.

Each spreadsheet had its own format, assumptions, and structure, making data exchange a nightmare. Something as simple as a column order change could break entire workflows, and reconciling information across departments was time-consuming, error-prone, and frustrating.

For years, the industry accepted this as "just how things work." But it doesn't have to be.

What if data could structure itself?

This is the power of AI-driven Linked Data transforming raw, unstructured information into fluid, connected intelligence.

Just as Web 1.0 connected computers and Web 2.0 connected people, Linked Data connects information, breaking down silos and enabling seamless data interaction. AI-powered LLMs take this a step further, abstracting away the cognitive and analytical effort needed to transform raw data into structured, meaningful insights.

Instead of spending hours wrangling spreadsheets and databases, investment firms, procurement teams, and financial institutions could query knowledge directly, making sense of disparate data in real time.

Converting Siloed Spreadsheets to Connected Data

The impact? Faster insights. Smarter decisions. No more data dead-ends.

Investment portfolios could update dynamically, pulling from live data rather than static files
Teams could ask questions and get real-time, structured answers rather than relying on outdated reports
Legacy systems wouldn't have to be painfully integrated; they'd evolve into something better, more innovative, and more connected
No more wasted time migrating all of your data into a closed system that is expensive, rigid and limited only by the progress of its engineering team

The reality is that structured data isn't meaningful data, but with AI-powered Linked Data, we can finally bridge that gap.

"What if every industry instantly had seamless, intelligent access to all of its knowledge?"

How much time could we save?

How many opportunities could we unlock?

How much innovation could we inspire?

The future isn't just about collecting more data but making the data we have work for us.

The era of knowledgeable, interconnected data starts now.