AI Low-Code RFI and Submittals in Construction: Tools, Workflows, and Governed Sign-Off

AI low-code for construction RFIs and submittals is the pairing of review tools that read specs and drawings with workflow platforms that route each item and record who approved it. Together, the two cut the reading and routing delay. AI builds the submittal register, drafts cited RFI responses, and flags non-compliant submittals.

Low-code tools such as Microsoft Power Apps add ball-in-court routing into Procore or Autodesk Forma (formerly Autodesk Construction Cloud). A licensed professional signs off on every response.

At a glance

• AI reads the spec book and generates the submittal register, citing each item’s section.
• AI drafts first-pass RFI responses naming the exact drawing sheet and spec section used.
• Low-code platforms build the routing around Procore and Autodesk Forma.
• A licensed architect, engineer, or project manager signs off on every response and approval.
• Workflows built on a platform IT governs keep a timestamped record of who approved each item.

On an AI low-code platform, a team describes the RFI log or submittal approval chain it needs. The platform produces a working first version, and people refine it visually. The construction job has two halves: reading documents, and moving them between people and systems. Construction-specific AI review tools do most of the reading today, and low-code platforms build the moving.

For a CIO or IT director, the trigger usually arrives as a request already in the queue. A field team has built a faster submittal tracker on tools IT never approved, and now it wants that tracker connected to Procore. The stakes are schedule and liability. This roundup covers four AI review tools and six low-code options for routing RFIs and submittals. It ends with the governance question: can anyone prove the human sign-off later?

Why do RFIs and submittals stall construction schedules?

Most of that RFI cost is labor spent reading. Someone searches drawings, specifications, meeting minutes, earlier RFIs, and prior submittals for an answer that usually already sits somewhere in the project record.

Submittals fail in a similar way. Reviewers return a share of packages on first pass, and each rejection restarts the cycle. The subcontractor revises, resubmits, and waits for another reviewer to read the same spec section from the top.

RFI delays and submittal rejections share one root cause, manual reading. That is the work AI review tools take over first.

How does AI power the RFI and submittal workflow?

AI powers the reading half of the RFI and submittal workflow. It builds the register of required submittals from the project specifications, drafts RFI responses that cite their sources, and checks each package against its governing spec section. Nomic AI, Constructable.ai, and Helonic each cover part of that reading. Nomic AI, for example, drafts RFI responses with citations to the spec sections and drawing sheets behind each answer.

• Submittal register generation: AI reads the project specifications, identifies which divisions require submittals, and creates register items that each cite their spec section.
• RFI first draft with citations: AI searches drawings, specs, meeting minutes, and earlier RFIs, then drafts a response that names the sheet number and section it relied on.
• Spec compliance cross-check: AI compares each requirement in the governing specification section against the shop drawings or product data in a package. It flags deviations in a compliance matrix.

The AI cannot tell which deviation the architect will accept, so each flag goes to a reviewer as a question to answer. The time saved comes from triage: reviewers check the flagged items instead of reading each package cover to cover.

AI review speeds up reading but does not route anything. The cited draft still has to reach the right reviewer, in sequence, with a record of what happened.

What role do low-code platforms play in RFIs and submittals?

Low-code platforms run the routing half of the RFI and submittal workflow. Teams use them to build the app that decides who gets each item next, how long that person has to respond, and where the answer lands in Procore or Autodesk Forma. Microsoft Power Apps is one example.

The working pattern has three steps:

1. Ingest: An RFI or submittal package arrives from Procore, Autodesk Forma, or email, and the workflow logs it against its spec section.
2. AI review: The document passes to a spec-checking tool, which returns a cited draft response or a compliance matrix.
3. Low-code routing: The workflow assigns ball-in-court and starts the response clock. It escalates when the clock slips and writes the decision back to the system of record.

Teams can connect these steps to Procore with Zapier, which offers RFI and submittal triggers and actions, or with Power Automate through a third-party Procore connector in preview.

Once an RFI or submittal workflow needs approval tiers, role-based permissions, and a retained record, it is an application. Choosing its platform becomes a governance decision.

Which AI and low-code tools handle RFIs and submittals?

The tools below each focus on one half. Construction-specific AI tools such as Nomic AI handle document reading. Low-code options such as Microsoft Power Apps handle routing and integration. Each entry gives the role of the tool in RFI and submittal work, then its G2 rating where one exists.

AI review tools that read specs and drawings

Nomic AI

Role in RFI and submittal work: Nomic AI processes submittal packages and cross-references them against project specifications and drawings. It flags deviations, substitutions, and missing information before the package goes to the design team.

Constructable.ai

Role in RFI and submittal work: The Submittal Pre-Check in Constructable.ai compares uploaded product data sheets against the project specifications line by line, and each submittal links to its spec section.

Datagrid

Role in RFI and submittal work: Datagrid, a Procore company, offers a Deep Search Agent that cross-references submittals and RFIs against requirements and highlights discrepancies.

Helonic

Role in RFI and submittal work: Helonic is an AI construction drawing analysis platform that checks every sheet in a drawing set for issues. A reviewer then pushes each issue to Procore or Autodesk Forma as an RFI (checked 23 October 2026).

Low-code platforms that build the workflow

VeilSun

Role in RFI and submittal work: VeilSun builds custom low-code apps, primarily on Quickbase, and lists Procore among the technologies it works with.

Kahua

Role in RFI and submittal work: Kahua is a configurable construction platform where teams design their own apps, and its RFIs app runs on a Kahua workflow that administrators configure.

Ratings: 32 G2 reviewers give Kahua an average of 4.4 stars (checked October 2026).

Microsoft Power Apps

Role in RFI and submittal work: Teams use Power Apps to build RFI and submittal routing and ball-in-court tracking. It sits in Microsoft Power Platform alongside Power Automate, which automates workflows.

Ratings: 569 G2 reviewers give Microsoft Power Apps an average of 4.3 stars (checked October 2026).

Mendix

Role in RFI and submittal work: Mendix is designed for a variety of app developers, from business experts with no coding background to technical programmers. It builds enterprise web and mobile applications such as an RFI or submittal tracker (checked 23 October 2026).

Ratings: 417 G2 reviewers give Mendix an average of 4.4 stars (checked October 2026).

OutSystems

Role in RFI and submittal work: OutSystems sells a platform to build enterprise apps and AI agents, such as an RFI or submittal tracker (checked 23 October 2026).

Ratings: 1,810 G2 reviewers give OutSystems an average of 4.5 stars (checked October 2026).

Kissflow

Role in RFI and submittal work: Kissflow is an AI-powered enterprise application platform on which business teams and IT build, automate, and govern work. It sits around systems of record rather than inside them.

Ratings: 590 G2 reviewers give Kissflow an average of 4.3 stars (checked October 2026).

How do the low-code options compare on who builds and ratings?

The table compares the six low-code options, which handle the routing half of the job.

Governed low-code: who builds the workflow, and who keeps control?

Shadow AI usually starts with a field team that is right about the problem. Business teams should build the RFI and submittal workflow, and your IT team should keep control of it on one governed platform.

The request reaching your queue tends to look the same. By the time it arrives, project drawings and specs have often gone into a consumer AI model outside your enterprise agreements, and the approval logic sits in an app your team would need its builder to explain.

RFIs are late, and the team found a faster way to read. The fix is a sanctioned path that moves just as fast. You can hold any RFI and submittal workflow to these criteria:

• Project documents go only to AI services under enterprise API terms, never to consumer accounts.
• Every app that routes an RFI or submittal runs on a platform IT governs, with access set by role.
• Approval rule changes, whether a person or AI made them, land in one audit log your team can read.
• The workflow wraps Procore, Autodesk Forma, SAP, and Oracle instead of copying their records into a new store.

Governance does not fix a badly designed approval chain: someone still decides who reviews which spec division.

Where Kissflow fits

In an RFI and submittal setup, Kissflow would route and record decisions while systems of record such as SAP and Oracle keep their data, and the document reading stays with tools such as Nomic AI. It handles the structured work at the edges of those systems and never replaces them. Production deployments of Kissflow include new store opening and retail lifecycle management, oil rig shutdown and turnaround management, and end-to-end procurement and financial consoles.

Measured against the criteria above: every action lands in an audit log with a timestamp and the acting user, and changes AI makes sit in that same log beside manual edits. AI Builder, generally available since May 2026, builds a working app from one prompt or an uploaded requirements document, refined in conversation, and produces blueprints a developer can open and read.

For the due diligence an IT director runs on a field team’s request, Kissflow holds SOC 1 Type II, SOC 2 Type II, SOC 3, ISO/IEC 27001, HIPAA, GDPR, and CCPA certifications. It has over 1,200 customers worldwide and was included in the Forrester AppGen and Low-Code Platforms Landscape, Q2 2026.

Governance is what lets the field team keep its speed without your team inheriting an app it cannot audit.

Human in the lead: who signs off on AI-drafted RFIs and submittals?

The licensed professional signs, and accountability does not move to the software. AI output is decision support: a cited draft, a flagged deviation, a suggested register line. The architect, engineer, or project manager of record decides what is issued.

The decision also has to hold up later, in a claim, a dispute, or an audit, as part of a defensible project record. That record needs four things to survive scrutiny:

• The AI output exactly as the reviewer received it, with its citations.
• Every edit the reviewer made before issuing.
• The identity and role of the person who approved, with a timestamp.
• The reason the item moved, whether that was approval, a return, or an escalation.

The question that separates platforms is whether the workflow layer captures those four things by default or leaves them to someone’s diligence.

Frequently asked questions

Are RFIs considered submittals?

No. An RFI is a question about the project record, answered from drawings, specifications, meeting minutes, and earlier RFIs. A submittal is a package of shop drawings or product data checked against its governing specification section. AI tools handle them differently: they draft cited RFI responses and cross-check submittals against spec.

Is there an AI that can create construction documents?

Yes, for documents assembled from the project record, such as a submittal register, a cited RFI response draft, or a compliance matrix. Drawings and specifications still come from the licensed design team, which signs and seals what is issued.

Should the RFI and submittal workflow live inside a construction platform or sit around Procore?

Each option puts the workflow in a different place. Kahua is a configurable construction platform, so RFI workflows live inside it. Power Apps, Mendix, and OutSystems build separate applications that sit around Procore or Autodesk Forma.

Book a demo to see how Kissflow puts RFI and submittal approvals into one audit log.