Your competitive advantage shouldn’t belong to your SaaS vendor.
Build your own software. Not your own software company.
thesis3 is the trusted partner for insurers, MGAs, brokers and others across the specialty insurance industry, helping them build, own and operate their own technology. Own what differentiates you. Rent what doesn’t. AI-native engineering, insurance depth, and the discipline to run it safely in production.
Today, you rent back
your own advantage.
Every platform you buy hands a software company your competitive advantage. You give up flexibility. You give away the economics. You help build their software business with your unique IP, and then you rent it back, forever. And instead of building what makes you different, you spend the relationship arguing over scope and change requests.
It does not have to work that way. AI has given insurers the ability to build and own their solutions without becoming a software company.
The promise of AI wrote a cheque that SaaS vendors don’t want you to cash.
Own what differentiates you.
Rent what doesn’t.
Building everything is not the answer. Rebuilding commodity technology is as wasteful as renting your differentiation. The real question is simpler than the market makes it.
Trust, in three parts.
Trust is the organising principle. Everything thesis3 does sits under one of three headings.
Trusted Services
The expertise to build with confidence. Insurance knowledge, engineering, architecture, AI, security and operations. We help you decide what to build, what to buy, and how to put it together, then prototype, productionise and protect what we build with you.
Trusted Components
Do not build what is already a commodity. Proven components and platforms for the things that create no advantage: core infrastructure, workflow, identity, data services, document processing, integration and observability. Rent them. Do not rebuild them.
Trusted Solutions
Own the technology that makes you different. Agentic Submissions and the line-of-business-deep Underwriting Workbench, purpose-built around your underwriting, distribution, claims and decisioning. Built for your business, run safely in production, and owned by you.
Insurance depth and AI,
in the same room.
Some teams know specialty insurance. Some teams know how to build AI-native software. Almost no one is genuinely deep in both. That overlap is where thesis3 lives, and it is where the value is.
Own the intelligence
that makes you different.
Agentic Submissions, a line-of-business-deep Underwriting Workbench and a Claims Workbench, plus everything around them. Purpose-built for your business, live in weeks, and owned by you. Agents do the work. Your experts make the decisions.
Across the underwriting lifecycle
Submissions and data extraction
Ingests emails, PDFs and SOVs. Structured, triaged and de-duplicated in seconds.
Underwriting workbench
One workspace per risk: rating, guidelines and referrals in a single view.
Agentic workflows
Agents execute the busywork, the checks, follow-ups and data pulls, not just suggest.
Quoting, binding and endorsements
Quote, bind cover and process mid-term changes without leaving the platform.
Policy admin and PAS integration
Two-way sync with your PAS. No re-keying, no reconciliation drift.
Aggregation and exposure management
Live portfolio roll-up of CAT, footprint and line-level accumulation as you write.
Risk intelligence and insights
Third-party and internal data surfaced at the point of decision.
Legacy modernisation
Wraps and retires legacy systems incrementally. No rip-and-replace risk.
Generic technology fails specialty insurance.
Marine is not property. Property is not casualty. A generic platform forces every line through the same workflow. We build each line around how it actually underwrites. A few examples of that depth:
Marine
- Vessel and cargo valuation
- Voyage vs annual structures
- Port and routing accumulation
- Sanctions and flag-state screening
Property
- Location and peril scoring
- CAT accumulation control
- Rebuild-cost modelling
- SOV and schedule ingestion
Political Violence and Terror
- Footprint and radius aggregation
- Territory triage rules
- Event-based analytics
- Sanctions and OFAC screening
Casualty
- Long-tail reserving logic
- Coverage-trigger definitions
- Cross-layer aggregation
- Social-inflation loading
The pushback.
Why it does not hold
When you decide to build your own technology, your SaaS vendor will push back. It comes down to two arguments, and neither survives a second look.
They will tell you their software is your competitive advantage.
But an advantage they also license to your rivals is no advantage at all. What sets you apart cannot be something every competitor can buy off the same shelf.
They will tell you that you cannot build and run it safely yourself.
If it is that valuable and that hard, it is strategic, and strategic capability is the last thing to hand to someone else. The hard part is production discipline, and that is exactly what a partner is for.
The closer you look at the pushback, the clearer the case for owning what makes you different.
AI-native speed.
Production-grade trust.
No trade-off.
Meridian is the governance layer beneath everything we build. It is how we move at AI-native speed while meeting the quality, security and auditability an insurance business runs on. We did not build it to sell it. We built it because production demanded it, and proved it over eighteen months before it touched a single client.
Domain, standards and constraints are loaded before a line is written, and engineers approve every output before it ships.
Every module is built to be understood, with a full audit trail, so accountability never depends on one person.
Line-of-business rules, data models and constraints are first-class inputs, not retrofitted after the fact.
Governed code strengthens over time, while ungoverned debt accumulates silently until it breaks.
Start small. Own it fast.
You do not commit to a platform or a multi-year transformation. You start with a short, two to four week prototype of the thing that matters most, and decide from there. Our interests are aligned behind your vision, not a licence renewal.
Prototype
A working prototype of the capability that matters most, real and in your hands in weeks, built by people who know specialty insurance.
Productionise
Hardened into secure, scalable, auditable production technology, governed by Meridian, ready to run your business on.
Protect
Operated, monitored, maintained and improved. Yours to keep, with us alongside for as long as you want us.
Easy to claim. Hard to copy.
The market is full of platforms and consultancies making similar promises. The difference is in what sits behind them.
Built AI-first from day one.
Not an AI feature added to a legacy platform. AI is in the architecture from the ground up, which shows in what the system can do and in how fast it keeps improving.
Faster and leaner, by structure.
A small, senior team, AI-native, with no legacy platforms to maintain and no investor margins to defend. That structure lets us pass speed and cost to you, and it is hard to match without starting over.
Practitioners, not consultants.
Built by insurance people, engineered by AI-native builders. Deep specialty-insurance domain knowledge and AI-native engineering in the same room. Most vendors have one or the other. The combination is where the value is.
We are not new to this.
For over a decade, this team has built production software for specialty insurance, including an award-winning platform that ran in live underwriting and was acquired as a strategic asset. thesis3 is its next evolution. The 3 is provenance, not punctuation.
The last decade belonged to SaaS vendors. The next decade should belong to insurers.
Your differentiation should not belong to your SaaS vendor. That is the whole reason we exist.
Technology you can trust. Economics you can understand. Ownership you can keep.
Own what differentiates you. Rent what doesn’t. That is the whole idea. And every year you wait, the insurers who moved pull further ahead.