Holonic FAQ

Question 1

What is Holonic?

Accepted Answer

Holonic is the deterministic source of truth for critical software. CodeIntent, Holonic's platform, derives evidence directly from a customer's actual source code, accounts for every artifact, traces what changed, and flags what needs human review before a modernized system moves toward production acceptance.

Question 2

What does CodeIntent do, in plain terms?

Accepted Answer

CodeIntent reads a legacy codebase and builds a source-derived model of what the system actually does, then regenerates that system into modern, idiomatic code while showing exactly what was preserved, changed, added, or flagged for review. Nothing is rewritten from a guess. The baseline is built from the code itself, not imposed on it.

Question 3

What industries does Holonic serve?

Accepted Answer

Holonic is built for organizations where a modernization mistake is expensive or dangerous. Banking and financial services, federal and defense programs, healthcare payers, and large enterprise IT all operate in environments where "the AI seems confident" is not an acceptable basis for a production decision.

Question 4

What are the six disposition states in CodeIntent?

Accepted Answer

Every artifact in the source system is placed into exactly one of six states: Verified, Preserved, Allowed Change, Net-New, Review Boundary, or Orphaned. Every artifact dispositioned, zero unexplained. Nothing disappears silently between the old system and the new one.

Question 5

How do you know nothing in the codebase was missed?

Accepted Answer

CodeIntent Studio covers the entire codebase, including undocumented and hard-to-read sections, and records every artifact and its status in the Scope Ledger. Where operational intent can't be determined, the element is surfaced in Governed Boundaries for a human decision rather than resolved silently. Every artifact dispositioned, zero unexplained.

Question 6

What does "deterministic" mean here, and why does it matter?

Accepted Answer

Deterministic means the same input produces the same, rule-governed output every time. Unlike a probabilistic AI model that can behave differently on the same code from one run to the next, a deterministic system can hold its accuracy at scale. That matters because a probabilistic tool tends to degrade as a codebase grows into the millions of lines; a rule-governed one does not.

Question 7

Does CodeIntent rewrite my code, or does it read it?

Accepted Answer

CodeIntent reads the source; it does not rewrite it into something unrecognizable and ask you to trust the result. The evidence model is derived from your actual codebase, so the baseline reflects what your system really does, not what a model assumes it should do.

Question 8

Which languages and target architectures does CodeIntent support today?

Accepted Answer

The most automated path today is COBOL to Java Spring Boot, proven on customer-scale corpora including federal and defense workloads. Additional language accelerators, including TAL and C/C++, and domain-specific compliance frameworks are on the roadmap for 2027. Check with your Holonic contact for the current state of a specific language pair.

Question 9

What happens when the source code is ambiguous?

Accepted Answer

When CodeIntent can't determine a single, defensible interpretation of the source, it fails explicitly instead of guessing. The element is surfaced in Governed Boundaries for a human decision rather than buried inside code that merely looks finished. Ambiguity is made visible, not hidden.

Question 10

What's the difference between "proof" and "evidence" in Holonic's approach?

Accepted Answer

Holonic keeps proof and evidence separate on purpose. Proof of accounting means every artifact in the finite source codebase is accounted for and nothing goes missing. Evidence of behavior means Holonic provides source-derived evidence that the modernized system behaves as intended and demonstrates equivalence. Proof of accounting. Evidence of behavior. Naming that boundary, rather than blurring it, is part of the credibility case.

Question 11

Can I inspect the evidence myself, or do I have to take Holonic's word for it?

Accepted Answer

You inspect the evidence yourself. The entire design point of a source-derived evidence layer is that a reviewer, auditor, or engineer can trace any output back to the exact source structure and rule that produced it, rather than accepting a vendor's claim of accuracy on faith.

Question 12

We currently rely on a multi-year parallel run to trust a system migration. Does CodeIntent replace that?

Accepted Answer

CodeIntent is built to shorten the distance between generated and production accepted by giving reviewers verifiable evidence. Testing should confirm behavior, not discover what the LLM code generator forgot. On supported patterns, that means governed modernization measured in weeks rather than years, with the evidence to accept it delivered alongside the code. What ships in that window is a verified, fully traceable equivalent baseline your team then iterates and hardens toward production, not a turnkey cutover.

Question 13

Is CodeIntent suitable for regulated industries like banking, healthcare, or defense?

Accepted Answer

Yes, regulated environments are exactly where Holonic is built to operate. Current and prior engagements include mission-critical embedded systems work for Raytheon Technologies and an assembly language to Java modernization for the IRS via Maximus. Regulators and legal analysts covering AI contracting have independently flagged the same problem Holonic addresses: non-deterministic AI output is difficult to accept as sufficient proof of production readiness on its own.

Question 14

Does using CodeIntent change our compliance or audit obligations?

Accepted Answer

CodeIntent does not replace your compliance program. It gives your team and your auditors a source-traceable record of what changed, what was preserved, and what still requires a human decision, which is the kind of evidence an audit or model risk review typically asks for. Talk to your Holonic contact about how the evidence layer maps to your specific framework, including SOX, model risk guidance, DO-330-style tool qualification, and similar programs.

Question 15

Can customers edit CodeIntent directly today?

Accepted Answer

Not yet. Today, CodeIntent Studio is a governed, inspectable, read-only baseline. The ability for a business owner to edit intent directly and generate against the new baseline is targeted for fall 2026 in V1.1.

Question 16

Does CodeIntent automatically block or reject code that doesn't match the baseline?

Accepted Answer

No, CodeIntent does not automatically block or reject code today. It holds, checks, and traces every element of the source system, and it fails explicitly and surfaces a required human decision when behavior cannot be determined, rather than guessing or silently accepting a change. Automated enforcement is a later stage of the roadmap, not a claim Holonic makes about the current product.

Question 17

Can CodeIntent add new functionality, or does it only convert what already exists?

Accepted Answer

Both. CodeIntent Studio carries existing behavior forward as Preserved or Verified, while also supporting Net-New functionality. New capability is explicitly identified, governed under the same baseline, and remains fully traceable to the decision that introduced it. Modernization and new feature development can happen in the same pass.

Question 18

What happens after go-live?

Accepted Answer

The CodeIntent baseline remains with the application after modernization. It becomes the governed baseline for future human- and AI-generated changes, helping reviewers understand what changed, what remained the same, and how proposed changes relate to the original operational intent. Drift is surfaced instead of accumulating silently.

Question 19

How long does a modernization project take?

Accepted Answer

On supported patterns like COBOL to Java, governed modernization runs in weeks rather than the years a traditional services engagement typically takes. The evidence to support production acceptance is delivered alongside the code. Timelines still depend on your specific system, access, and internal approval process, so treat this as a general expectation rather than a fixed date.

Question 20

Who typically works with Holonic: our internal team, or a systems integrator?

Accepted Answer

Both direct and systems-integrator models work. Some organizations bring Holonic in directly alongside their internal engineering and compliance teams. Others engage through a systems integrator partner already running their modernization program. Either way, CodeIntent's evidence layer gives everyone in the room, including internal engineers, external SIs, and auditors, a shared, inspectable account of what changed.

Question 21

Can our developers keep using their existing AI coding tools?

Accepted Answer

Yes. CodeIntent is built to work alongside AI coding tools such as Cursor, GitHub Copilot, and Claude Code, not replace them. Your developers keep generating code however they already do, and CodeIntent checks every proposed change, human-written or AI-generated, against the governed baseline before it's accepted. The tools propose; CodeIntent governs.

Question 22

How do we get started?

Accepted Answer

Start by scheduling a demo or working session with Holonic. The team will review your modernization goals, source-system profile, governance requirements, and evidence needs, then identify whether CodeIntent is a fit for the first supported path.