Built on forty years, ready for what comes next: A story of specialty insurer's modernization with Symfa.

Our client is a specialty insurer with more than forty years of operating history. Deeply established in their segment — credit insurance and loan-protection products distributed through banks, credit unions, auto dealers, and consumer finance companies. An excellent financial strength rating from AM Best. Institutional knowledge of this corner of the insurance market that is, by any measure, exceptional.

When an organization of that standing decides it is time to evolve, it means something. It is not a reactive decision. It is a deliberate choice by leadership that has seen enough market cycles to know exactly when to invest in the next era — and exactly who to trust with the work.

We kicked off in September 2025. Within two months, 14 Symfa specialists were on the account across documentation, modernization, platform development, and data engineering — all running in parallel within a live operation that kept moving throughout.

Here is what those 32 people are actually building.

The first and most technically demanding initiative involves what's known in the industry as an AS400 platform.

AS400 is an IBM environment built to run RPG, a programming language that predates much of modern software development. In the right hands — as it has been in our client's operation — it is robust, reliable, and highly efficient. In their case, it has processed their core debt protection operations faithfully for decades. Every certificate sold through their distribution network is validated, calculated, and confirmed through this system.

The challenge was not that it had failed. It hadn't. The challenge was that the knowledge required to interpret, document, and eventually modernize it exists in a form that very few engineers can access. RPG code at this level of maturity and density is not readable through standard tooling. Modern AI systems cannot reliably interpret it. What's required is an engineer with the specific technical background to read low-level code the way a specialist reads a specialist's notation — the kind of expertise you either have or you don't.

We had that person. One of our engineers, with a background in assembler and low-level systems, spent weeks working through the platform's modules — extracting the real business logic and building a comprehensive written record of a system that had been operating on institutional expertise alone.

The result is something more durable than a headline: a complete, accurate understanding of a core system, documented in full for the first time. The modernization comes next — but it comes from a position of knowledge rather than assumption.

This is what we do when the brief is genuinely hard. We find the person who can actually read it.

The agent services workstream is a different kind of challenge — less about technical archaeology, more about translating years of refined human practice into structured, scalable software.

Our client's insurance products are distributed through agents: banks, auto dealers, credit unions, and consumer finance organizations. The process of onboarding a new agent — establishing their account, making them operational in the system — was carried out by a team that knew exactly what they were doing. This was work shaped by years of experience, executed by people who understood every nuance of what the business required.

The goal was never to replace that expertise. The goal was to honour it — to take a process refined over many years and give it an architecture that could scale without relying on the same small group of individuals to hold all of it in memory.

We started with observation and deep conversation. Interviews with the people doing the work. Screen recordings. Live walkthroughs. We wanted to understand not just the official steps, but the real ones — including the judgment calls that experienced operators make instinctively and that no formal documentation had ever captured.

By May 2026, the result of that process was entering user acceptance testing: an external portal for agent representatives, an internal portal for operations staff. Structured. Auditable. Built to scale.

What makes this kind of work satisfying is that it honours the expertise it's built around. The people who held this process together did something genuinely skilled. Our job was to ensure that skill outlasts any single person's tenure in the organization.

Data Warehouse — Azure Lakehouse

Before: Financial and operational reporting was assembled by hand. Analysts pulled figures from multiple sources — Power BI reports, legacy Access outputs, manual spreadsheets — reconciled them manually, and entered the results into the accounting platform. Month-end close was a multi-day, labour-intensive process with meaningful exposure to manual error.

After: A full Microsoft Azure Lakehouse: ADLS Gen2 storage, Azure Synapse Analytics, a Bronze-Silver-Gold data pipeline that cleans and structures incoming data automatically, and Power BI dashboards that reflect the latest numbers without manual assembly. The accounting platform receives the data it needs directly.

Where we are: The foundational pipeline is live and reconciled across hundreds of thousands of transactions. The month-end close automation is currently in the implementation phases.

What comes next: Once the closing process is fully migrated, the scope expands. The client's ambition is a full transition to the cloud — moving all operational data and reporting infrastructure off legacy systems entirely. The Lakehouse being built today is the foundation that makes that possible.

Business value: Reporting cycles compressed from days to near-real-time. Month-end close transformed from a reconciliation marathon into a verification step. Foundation established for future analytics, automation, and AI applications across the business.

Security Compliance Documentation — Completed

Before: The organization had no centralized system inventory, no formal process maps, and no documentation of how systems from different departments interact. The knowledge existed — held by experienced people across the organization — but not in any form that could survive personnel changes, support audits, or be used to plan infrastructure work with confidence.

After: A structured documentation program built department by department: full application inventories, workflow diagrams, data flow maps, and cross-department dependency records. Each department's systems documented from first principles, with diagrams showing how everything connects.

Business value: Level 3 compliance maturity within reach. A single source of truth for IT governance established for the first time. Foundation laid for SOC2, HIPAA, and ISO audit readiness. The organization can now plan infrastructure changes against an accurate map of what they have.

Certificate Processing Automation (TransCredit)

Before: Monthly processing of certificates — validating and reconciling the products sold through agent channels — was managed through Access and Excel. Accurate, but slow; corrections were manual; month-end balance required significant effort from the team managing it.

After: A unified web module within the platform handles the full certificate lifecycle — issuance, corrections, cancellations, claim refunds, agent payment reporting, and batch intake — replacing a network of Access databases and manual workflows that had grown over many years. Full audit trail at every step.

Where we are: Core functionality is developed and we are ready for UAT. Final implementation phases are underway — covering reporting, batch processing completion, and the Admin tab — with delivery expected within the next four to six weeks.

Business value: Month-end balance cycle accelerated. Error correction made traceable and structured — every change logged, every reason recorded. Agent reporting consolidated into a single visible system. And a process that previously depended on a small group of people holding years of institutional knowledge now runs on documented, auditable workflows available to the whole team.

Microsoft Dynamics Implementation

Before: Sales and account management ran across Outlook, Excel, and a basic internal CRM. Client contacts lived in email. Reporting meant pulling data from several places and assembling it by hand. No unified view of relationships across the business.

After: Microsoft Dynamics 365 Sales configured as the single platform: standardized pipelines, complete client records, integrated reporting, a 360-degree view of every account.

Business value: Full visibility across client relationships in one place. Sales activity trackable without manual assembly. CRM infrastructure ready to scale with the business.

Claims MVP

Before: Agents filed claims through paper-based or phone-based channels. Internal staff received and routed them through manual intake.

After: A digital filing portal for agents; an internal review portal where operations staff can approve, deny, or request additional documentation on each submission. End-to-end digitised.

Business value: Claims intake modernized. Agent experience improved. Internal team able to manage volume through a structured queue with full visibility into each case.

Accounting Automation — Journal Entry Integration

Before: Month-end close was a multi-day manual process. Source data was gathered by hand from MS Access, Power BI, Excel spreadsheets, and multiple other systems, consolidated in Excel separately for each legal entity, then exported as text files and manually uploaded into Microsoft Dynamics GP via SmartConnect. The process was time-consuming, error-prone, and dependent on institutional knowledge held by a small number of people.

After: 46 journal entry integrations — covering underwriting premiums, claims, commissions, management fees, premium tax, reinsurance, and more — now run directly from SQL to Microsoft Dynamics GP via SmartConnect, with no Excel staging and no manual file uploads. Each integration includes a pre-run preview showing calculated debits and credits before anything is posted, transaction rollbacks on failure, and centralized logging via OpsLog. The accounting team opens SmartConnect, selects the month, and clicks run.

Business value: Manual calculation and file preparation eliminated across all journal entry types spanning multiple lines of business. Month-end close reduced from a multi-day reconciliation effort to a supervised, one-click process. Full audit trail established — every entry traceable to its source data and logic. Architecture designed to accommodate frequent re-closings and multiple open months simultaneously, reflecting the operational reality of insurance accounting

Vendor ID Management

Before: Vendor records in Microsoft Dynamics GP were created without a centralized validation layer. The same agent or vendor could exist under multiple IDs across different systems, leading to duplicate records, misrouted payments, and downstream reconciliation failures.

After: A controlled provisioning pipeline connected to the Agent Services platform. New vendors are validated, deduplicated, and created through a structured flow — with EntityMaster as the single source of truth and clear ownership of the vendor lifecycle.

Business value: Duplicate vendor creation eliminated at the point of entry. Reconciliation errors reduced. A single, auditable source of truth for vendor identity established across all systems.

Ten workstreams. 32 people. The work is varied in type and complexity, but it shares a common thread: this is an organization that made a deliberate decision to modernize from a position of strength — not under pressure, not reactively, but with a clear view of the business it is building toward. Their leadership saw what needed to evolve before it became urgent. That kind of foresight is what separates organizations that lead their categories from those that spend years catching up.

What comes with that kind of partnership is a specific kind of accountability. When our client's senior counterpart told us in that conference room that he wanted to know if anyone on his side was creating friction for us, he was doing something important. He was making us responsible in both directions. He had backed us fully. He expected us to act accordingly.

That doesn't make the work easier. It makes it more serious. The margin for the comfortable explanation disappears. And what rises in its place is a different quality of effort — one where the ceiling on what you're willing to attempt moves upward because someone has genuinely staked their confidence in you.

We feel that every day. We are grateful for it.

The product development stream is the most forward-looking signal in the current portfolio. A post-MVP backlog is already taking shape. The immediate mandate was to help the client move beyond Access-based workflows — but the appetite for what comes next extends well beyond that starting point. There is a substantial roadmap of automation and optimization work ahead, and it is growing. 

A Symfa business analyst is now embedded in the process by which the client develops new insurance products — upstream from everything we've built so far. We are still learning the full landscape there. But we are in the room.

A second in-person visit is planned. The goal is to finally reach the client's primary offices — to meet the full operations team face to face, to understand the business through proximity rather than documentation alone.

We are looking forward to it. Not for the destination. For what it confirms: that this is a relationship built to continue.

We publish engineering stories from real projects — no sanitized case studies, no vendor-speak. If that's useful to you, follow Symfa on LinkedIn or come back for the next one. More on insurance modernization, legacy system transformation, and what embedded delivery actually looks like in the related articles below.

Follow Julia on LinkedIn for more on delivery management in insurance modernization.