Data migration doesn't break your data. It shows you how fragile it already was – and has been for years. However, what can break everything else – the timeline, the budget, the team – is underestimating what you're actually doing. Data migration shouldn’t just be a “line item in the project plan”. It's the continuos and iterative work of getting your data right so your business can operate right.
Data migration shows up in every program whether it is customer onboarding, system replacement, a modernization initiative, or an M&A integration – and it is always messier than anyone expects.
Data migration is consistently the highest-risk, most time-consuming activity in any systems change. And the reasons it goes sideways are remarkably predictable – even if teams keep getting surprised by them.
After years of working with financial institutions, consulting firms, and software companies on this exact problem, I've seen the same four patterns show up again and again. Understanding them is half the battle. The other half is knowing what it takes to get ahead of each one – the right approach, the right tooling, and the right mindset – before they compound into something program-threatening.
Every Production System Carries Operational Debt
People talk about technical debt in code. But production systems carry something broader: business operational debt. Years of workarounds, bolt-ons, manual overrides, and undocumented exceptions that kept the business running. When you migrate, that debt doesn’t stay behind. It shows up as data – messy, inconsistent, and full of edge cases nobody remembers creating.
This is why upfront and ongoing data profiling is critical at the start and throughout any migration. When you can see the completeness, distribution, and quality of your data within minutes rather than weeks, you’re working from reality instead of assumptions. A project manager who knows upfront that a critical date field is missing in 500 records can plan around it. One who discovers this for the first time three months in is managing a crisis.
The Problem Lives in the Handoffs
Here’s something I see on every program: the person who knows the business rule is not the person who writes the data rule. Between them, there’s a chain of handoffs – analysts, engineers, sometimes third-party consultants – and every stop is a lossy connection. Context gets dropped. Intent gets reinterpreted. By the time a transformation rule gets coded, it may reflect what someone thought the requirement was, not what it actually was.
The compounding effect is brutal. One misunderstood business rule becomes a transformation error, which becomes a reconciliation break, which becomes a go-live delay. If the person who knows the answer could act on it directly – without the chain of handoffs – most of these breaks never happen.
Most Programs Start from the Wrong End
It's worth separating two things that often get conflated: lift-and-shift and data migration. Lift-and-shift is moving or replicating data without logical change to data. A true data migration is something different. It's an opportunity to land in a target state – often with a data model change – that supports how the business operates going forward, not how it operated before.
That distinction changes where you should start. The typical instinct is to start with what you have: pull out the source data, understand it, and then figure out where it goes. That feels logical. But starting from the source means you can invest significant effort in mapping and transformation before you fully understand what the target actually requires. Gaps appear slowly – or worse, after significant work has already been done.
A target-centric approach flips this. Start with what the new system requires, then work backward to understand how your current data fits – or doesn’t. AI-powered mapping can predict field matches between source and target schemas in seconds, giving teams a starting point that would otherwise take days, weeks or months of manual side-by-side comparison. That head start changes the trajectory of the entire program.
In Financial Services, Complexity Is Structural
Not all data migrations are created equal. When you’re migrating investment or financial applications, the complexity isn’t just about volume – it’s structural. Financial data doesn’t live in one place. Positions, counterparties, reference data, and transactions are scattered across systems, each with their own rules, formats, and interdependencies.
At this level of referential complexity, you need more than a mapping spreadsheet. You need metadata that actively connects every migration step – so when one field changes, everyone downstream knows about it. And if you’re dealing with legacy mainframe systems, the challenge compounds further: the business logic that governs how data was calculated, stored, and routed is buried in COBOL modules that may not have been documented in decades.
How Zengines Helps You Get Ahead to Avoid the Mess
Data migration isn’t a side activity that happens at the end of a program. It’s the connective tissue of every systems change – whether you’re modernizing legacy systems, managing mainframes, or meeting new regulatory compliance requirements. We built Zengines to treat it that way.
Every problem I described above has a direct answer in our platform.
- Operational debt hiding in your data? Zengines profiles your source data automatically – surfacing completeness gaps, format inconsistencies, and quality issues in minutes instead of weeks, so your team plans from reality, not assumptions.
- Challenging handoffs between business and technical teams? Our platform keeps analysis, mapping, transformation, and reconciliation in one place, so the person who knows the business rule can act on it directly – no chain of handoffs, no lost context.
- Starting from the wrong end? Zengines is target-centric by design: AI predicts field mappings between your source and target schemas in seconds, giving teams a validated starting point that would otherwise take days of manual comparison. AI also generates transformation rules to ensure the data gets the right business logic treatment.
- And the structural complexity of financial data? Our platform maintains active metadata that connects every migration step, so changes upstream are visible downstream – across every table, every relationship, and every transformation rule.
When legacy mainframes are part of the equation, Zengines goes further. Our contextual data lineage capability parses COBOL, RPG, and PL/1 code to extract the embedded business logic, calculation rules, and data flows that have been locked inside these systems for decades – giving your team the transparency to reverse-engineer requirements in minutes, not months.
The result: business analysts are 6x more productive, migrations move 80% faster, and transformation rules are generated from plain English prompts – so the people closest to the business drive the process without waiting on engineering resources.
The programs that go smoothly aren’t the ones with the simplest data. They’re the ones that saw the potential messiness early, connected the right people to the right decisions, and had the tooling to act on what they found.
If your organization is planning a migration or modernization initiative, schedule a demo with our team to see how Zengines turns the messiest part of your program into the most predictable one.
You may also like
At industry conferences this year, I’ve spent dozens of hours inside conversations with CEOs, CDOs, CIOs and operating executives across financial services. When I ask what’s keeping them up at night when it comes to their data, the answer is remarkably consistent: data access. They want data more accessible, faster, in more usable form, in more places, with fewer gatekeepers.
What's notable is what they don't ask for. Not trustworthiness. Not audit-ability. Not the ability to defend a number to a regulator without calling three people first. Access is the ceiling of the conversation, and honestly, that makes sense. In large financial enterprises built on decades of legacy applications, murky integrations, and pipelines that nobody fully documented, just getting the data somewhere useful is still a meaningful achievement.
The problem is that "getting the data" is already more complicated than most leaders realize. The moment data leaves its source system, decisions are being made about it. Decisions that quietly change what it means. And if you don't know those decisions were made, you don't know what you're actually looking at.
That's where lineage comes in, and why it matters even before you get to the outcomes leaders should be asking for.
Below, I’ll walk through (1) what “access” really delivers, (2) the abstraction layer hidden inside every extraction, (3) the compounding problem of “data derivatives”, (4) a concrete example – encoding and precision – where this gets expensive, and (5) what business leaders should be asking for instead.
What “Data Access” Really Delivers
When a business team asks for access to data, they almost always receive something that has already been processed for their consumption. Someone – usually a data engineer or database administrator – sat down with the source system and made a series of decisions:
- Which tables matter for this use case
- Which fields to expose
- How to filter, aggregate, or join the records
- Which technical artifacts to strip out (temp tables, system metrics, audit fields that don’t translate to business meaning)
These decisions are reasonable. Business consumers don’t want raw operational data; they want something readable without extraneous noise. But every one of those decisions encodes logic and judgment that doesn’t travel with the data. The output looks complete – and to the business user, it looks like the source of truth – but it is already an abstraction.
The Extraction Event Is a Translation Event
I find it useful to think of an extraction as a translation. Someone translated the operational reality of a data storage system into a business-readable view. Like any translation, choices were made: what to keep, what to drop, how to render concepts that don’t map cleanly across contexts. And like any translation, those choices can quietly change the meaning.
When a business leader looks at the extracted view, the assumption is usually that the data was “moved and shifted” – that is, copied with fidelity. That assumption is possible. In my experience, it is also highly doubtful. Logic gets applied at the moment of extraction, and unless someone deliberately captured and shared that logic, it is invisible by the time the data reaches a dashboard.
Abstractions of Abstractions: How Data Derivatives Compound the Problem
Here is where it gets harder.
Once an extracted data set exists, other people start using it. And why wouldn't they? There is already a data access path. The alternative - forging a new data access path - is the full corporate yellow tape headache: hunting for a charge code, filling out a technical work request that Business can’t quite decipher, watching that ticket age in a queue, and depending on legacy data SMEs who left the company in 2019. The extracted data set skips all of that. Already shaped for consumption, already lightly documented, already trusted by some peer team who vouched for it in a meeting six months ago. So the next team builds a report off it. Or creates a derivative data set for their own use case. Or both. What they don't realize is that the easy path and the right path may not be the same one.
They use it because it’s available and easier than starting from scratch – it’s already shaped for consumption, already lightly documented, already trusted by some peer team. So they build a new report off it. Or they create a derivative data set for their own use case. Or both.
That derivative is now an abstraction of an abstraction. The further you move from the originating system, the more layers of unrecorded judgment sit between the business decision and the operational event the data was supposed to describe. By the third or fourth hop, the question “where did this number come from?” can be genuinely difficult to answer – even for the team that produced the report.
A Concrete Example: How Encoding and Precision Quietly Rewrite Your Data
Let me make this concrete with an example I keep encountering.
When data is moved between systems, engineers make practical choices about how to package it. One of those choices is how to handle numeric precision. A value originally stored at six decimal places in the source might be packaged at four, or two, depending on what the receiving system supports – or simply what the engineer is most familiar with.
In some industries, that’s fine. In financial services, insurance, and healthcare, it is often not fine. A decimal place in an interest rate, a reserve calculation, or a pricing model can represent material variance. Once precision has been silently reduced, the data is no longer the real data – it is an approximation that looks identical to a casual reviewer. The business consumer assumes they’re working with the underlying record; in reality, they’re working with a rounded version of it that was reshaped during packaging.
This is exactly the kind of change that lineage is built to surface. Without lineage, you can’t tell that anything happened. With lineage, the precision change is documented, traceable, and reviewable.
Why Regulated Industries Can’t Afford to Skip Data Lineage
Regulatory frameworks have been ahead of business intuition on this point. BCBS-239 requires banks to demonstrate the accuracy, completeness, and timeliness of their risk data – which is impossible to defend without lineage. ORSA and Solvency II require insurers to substantiate the data flowing into solvency and capital calculations. None of these frameworks ask whether you have access to the data. They ask whether you can prove what the data is and how it got there.
For institutions operating under these regimes, lineage isn’t a nice-to-have analytics enhancement. It is the substrate that makes the rest of the data conversation defensible.
What Business Leaders Should Be Asking For Instead
If “give me access to the data” is the wrong ask on its own, what’s the right one? In my view, business leaders should be asking three questions every time a new data set lands on their desk:
- Where did this data originate, and what happened to it between then and now? Not a verbal summary – a documented path that is understandable in Business terms.
- What decisions were made during extraction or packaging that could have changed the meaning of the values I’m looking at? Especially around encoding, precision, filtering, and aggregation.
- If a regulator or auditor asked me to defend this number tomorrow, do I have the evidence trail to do it? If the answer is “we’d have to go find the engineer who built this,” the answer is no.
These questions don’t replace the access conversation – they extend it. Access is the entry point. Lineage is what makes access trustworthy.
A Final Thought
The reason business teams don’t ask for lineage isn’t that lineage doesn’t matter. It’s that the absence of lineage rarely announces itself. The data looks fine. The dashboard renders. The report mostly ties out. The risk lives in the assumptions you didn’t know you were making about what the data went through to get to you.
If your business teams are only asking for access, you have a gap – and in legacy environments where decades of undocumented logic sit between the source and the report, that gap is widest. The fix is to start asking for lineage too.
See Contextual Data Lineage in Action
Zengines Contextual Data Lineage is built for the environments where the lineage gap is widest – large financial enterprises with critical business logic locked inside COBOL, RPG, PL/1, and AS/400 code. We extract that embedded logic, make the data path visible, and give your teams the evidence trail they need to defend their numbers to auditors, regulators, and themselves.
If you’re working through a BCBS-239, ORSA, or Solvency II mandate, a planned mainframe migration, or a growing trust gap between your business teams and the data they consume, we’d like to hear about it.
In 2006, British mathematician Clive Humby coined a phrase that would define the next two decades of enterprise thinking: "data is the new oil." A decade later, in May 2017, The Economist made it a cover story – declaring data the world's most valuable resource and arguing that the data economy demanded a new approach to competition itself.
Twenty years after Humby first said it, the metaphor has only become more apt. What's changed is the catalyst. AI – and specifically the broad accessibility of large language models – has turned the abstract value of data into something organizations can now act on, at scale, in their actual operations. Every enterprise executive and Board member conversation I'm in today centers on the same question: are we positioned to scale value from AI?
The honest answer for most financial services enterprises is: not yet. And the gap isn't model selection, infrastructure, or use case prioritization. The gap is data readiness.
This post lays out what "AI-ready data" actually means in an enterprise context and the two capabilities that determine whether you have it.
What "AI-Ready Data" Actually Means
Strip away the hype, and AI-ready data comes down to two things:
- The data has to be available – meaning it can be moved, accessed, and used by modern systems regardless of where it originally lived.
- The data has to be trustworthy – meaning you know and can explain what it is, where it came from, and what business logic shaped it.
Both sound obvious. Neither is easy. And in older institutions with legacy applications – like in financial services – where institutions are sitting on decades of data stored across generations of systems, both require deliberate enterprise capability.
Pillar 1: Data Usability
Decades of preserved data only retains its value if the organization can keep it working. That means the ability to move it, transform it, and deliver it in a form whatever comes next can ingest; a new platform, a new analytics layer, an AI tool. Without that organizational capability, preserved data becomes stranded data.
Making data persistently usable across system changes is a data migration problem.
For institutions that have spent decades preserving customer records, transaction histories, account positions, and policy data, that preservation only translates into value if the data remains usable today. Not in the form it was stored in 30 years ago. In the form your current systems, your current analysts, and your current AI tools can ingest.
That's where data migration comes in – and where I'd encourage every executive to reframe how they think about it.
For most of the last 20 years, data migration has been treated as a one-time, project-bound activity tied to a specific systems initiative. A core conversion. A CRM rollout. An acquisition. A means to an end – the job had a start date and an end date, and once the data was "moved," the team and tools were disbanded.
That framing made sense in a world where systems changed every 10 to 15 years. It doesn't make sense anymore. The pace of modernization – driven by cloud adoption, AI tooling, vendor consolidation, and M&A – means data is constantly in motion. Treating each move as a bespoke, manually-staffed project is what makes modernization slow, expensive, and risky.
We built Zengines' data migration platform on a different premise: that data migration is a change capability, not a one-time activity. It's how you ensure your data remains an asset across every system change you'll make in the next 20 years – regardless of source format, target schema, or technology stack. That's what makes the underlying asset AI-ready: portable, repeatable, accessible.
For ISVs, BPOs, and MSPs onboarding clients onto modern platforms, the same logic applies and the economics are even more direct. Data conversion is, as I've argued before, a CEO-level concern – every client conversion that takes six months instead of six weeks is revenue deferred. Our platform compresses onboarding timelines by up to 80% by automating the manual work of mapping, profiling, transforming, and moving.
Pillar 2: Data Trustworthiness
Trustworthiness has many dimensions; data quality, governance, compliance controls. But none of those can be properly established without first answering a more fundamental question: what does this data actually represent, what logic produced it, where did it come from, and why does it look the way it does? That's a lineage problem, and it has to be solved before the rest can follow. In legacy-heavy environments, it's even harder to answer.
Trustworthiness matters on two distinct fronts:
First, the consumers of AI outputs; analysts, risk managers, portfolio teams; will act on what they trust. AI outputs will certainly attract interest; but that confidence erodes the moment someone is in a hot seat and can't explain a result, defend a decision, or reconcile an inconsistency. Without traceable source logic, that moment is a matter of when, not if.
Second, regulators are already examining AI model inputs. Under regulatory frameworks like BCBS 239, ORSA, Solvency II, "we trained on legacy system output" is not an explanation. The explanation lives in the code.
This is where data lineage matters, and where financial services has a particular challenge.
A significant portion of the data that drives banking, insurance, and asset management still flows through legacy systems – mainframes and the codebases that sit on them: COBOL, RPG, PL/1, Assembler. These systems weren't built to expose their logic to outside observers. The data they produce reflects calculations, conditional branches, and business rules that were written decades ago, often by people who have long since retired. When a CDO asks today, why does our risk exposure calculation produce this number?, the answer is buried in code that no current analyst can quickly read end-to-end.
At one Fortune 100 financial institution we work with, the environment includes nearly 100,000 COBOL modules. That's not unusual for an enterprise of that scale. It's the norm.
Without a way to expose the logic embedded in those systems, AI initiatives that touch this data are flying blind. You can train a model on the outputs, but you can't explain the outputs. You can move the data, but you can't verify what it represents. For regulated institutions, that's a non-starter.
This is the problem Zengines' Contextual Data Lineage solves. It parses legacy code – COBOL, RPG, PL/1 – and surfaces the business logic embedded inside: calculations, branching conditions, data origins, downstream dependencies. Instead of waiting nine months for a subject matter expert to reverse-engineer a single business rule, an analyst can answer the question in minutes. That's what makes legacy data not just movable, but explainable. And explainability is what makes data AI-ready in a regulated environment.
Why This Matters Now
The institutions making the most progress on AI right now aren't the ones with the most ambitious model strategies. They're the ones who've done the unglamorous work on the foundation – ensuring their data is preserved across system changes, and that the logic embedded in their legacy systems is documented, understandable, and ready to be replicated or retired with confidence.
That foundation is what allows AI initiatives to move from pilot to production to scaled value. It's what allows risk teams to validate AI-driven outputs against regulatory expectations with confidence. It's what allows finance and operations teams to actually trust what AI is telling them.
The window to build this foundation is now. Every quarter spent treating data migration as a project – or treating legacy code as an unsolvable black box – is a quarter of AI value deferred.
Two Capabilities, One Outcome
AI-ready data isn't a destination. It's the natural outcome of two capabilities working together: the ability to move data through any transformation or modernization without losing it, and the ability to understand the logic that defines what the data means over time and pathways.
Zengines was built to deliver both. Our data migration platform makes data preservation and utility a repeatable, AI-accelerated capability. Our Contextual Data Lineage exposes the logic locked inside legacy systems so analysts, auditors, and AI tools can use it with confidence.
If your organization is wrestling with how to position your data for AI – whether that's preserving decades of records through modernization, or making your legacy systems explainable to your CDO, CRO, or your regulators – we should talk.
See how Zengines accelerates the path to AI-ready data.
.png)
BOSTON, MA - May 8, 2026 - Zengines, Inc. today announced it has won Best of Show at FinovateSpring 2026, selected by audience and judges vote at the premier fintech demo event. The conference brought together more than 1,200 senior-level fintech and financial services executives - including 600+ from banks, credit unions, and financial institutions - to evaluate 50+ live product demonstrations.
Finovate recognized Zengines for its Contextual Data Lineage solution, citing the platform for "modernizing off mainframes without losing critical logic, satisfying auditors faster, and making legacy systems searchable so transformation and compliance don't stall."
Why it matters
Every financial institution running COBOL, RPG, or PL/1 has the same problem: the people who built those systems are retiring, regulators are asking questions the systems can't answer, and no one knows what a modernization program will actually touch until it's too late.
Zengines changes what's possible. Ask a plain-English question about your data. Get a complete, sourced answer - grounded in the actual logic embedded in the code, not a guess. Regulatory questions that took months get resolved in days. Migration risk gets quantified before work begins, not after.
Zengines is already working with a Fortune 100 financial institutions to navigate applications written in COBOL and RPG, each with more than tens of thousands of COBOL modules, cutting analysis time to minutes rather than months of manual research methods.
"Legacy system modernization has traditionally required a leap of faith - guessing what's in the code before you start rewriting it. We don't accept that. Contextual data lineage replaces guesswork with answers: regulatory questions resolved in days, business logic preserved through migration, and compliance that doesn't hinge on institutional memory. We're proving there is a better way to manage today and modernize tomorrow." - Caitlyn Truong, CEO and Co-Founder, Zengines
Watch the demo replay
About FinovateSpring 2026
FinovateSpring is the US West Coast's premier fintech showcase, bringing together innovators and banking decision-makers to shape the future of financial services. Best of Show awards are determined entirely by audience vote, with attendees rating companies on demo quality and potential impact.
About Zengines
Founded in 2020, Zengines is an AI-powered platform purpose-built for financial services data lineage and migration. The company helps financial institutions understand what is actually inside their legacy systems - so they can satisfy regulators, manage operational risk, and modernize without guesswork. Learn more or request a demo.
Subscribe to our Insights
.png)
Accelerate and de-risk data projects by 80%.
