Nutrient, the document intelligence company for modern businesses, announced a major update to its AI Assistant, adding a document editing agent that autonomously plans, executes, and adapts across multi-step tasks. The updated AI Assistant has access to Nutrient’s purpose-built document tools for rendering, structure-aware extraction, form operations, annotation, and redaction to complete end-to-end document workflows inside any application.
Enterprise teams evaluating document agents now face a build-versus-embed decision. They can assemble their own stack by stitching together LLMs, document parsing and rendering, tool orchestration, and governance controls, then maintain it as workflows and compliance needs evolve. Or they can embed a pre-built, policy-governed agent designed for production document work from day one. First introduced in early 2025, Nutrient AI Assistant is deployed directly inside the Nutrient Viewer SDK users already rely on, where it benefits from high-fidelity document rendering and purpose-built document tooling rather than brittle, one-off integrations.
With this release, Nutrient introduces an agentic document editing agent to AI Assistant. This new reasoning layer is able to plan and execute multi-step workflows using Nutrient’s purpose-built document tools (rendering, structure-aware extraction, form operations, annotation, and redaction), while being governed by custom skills — reusable task definitions that encode domain-specific logic — and three-tier approval policies (autonomous, confirmation-required, or prohibited). Developers don’t have to stitch together parsing, chunking strategies, prompts, and brittle tool calls. They simply embed the SDK, configure rules and permissions, and get a document agent that operates with production-grade accuracy and compliance inside their product.
Also Read: Gamma Launches Gamma Imagine to Bring AI-Native Design to the Masses
For document-heavy organizations, this means an agent that interprets what employees need, applies pre-configured skills and business logic, and produces results tailored to how they actually operate. For the developers building those applications, it means an embeddable SDK with an autonomous editing agent they can configure with their company’s rules, workflows, and compliance requirements — connected to the LLM provider of their choice.
“Every enterprise we talk to is trying to solve the same problem: they need AI that works with their documents, not just reads them. But they also need it to run inside their own application, under their own rules,” said Jonathan Rhyne, co-founder and CEO of Nutrient. “We’re giving developers and the enterprises they serve a document agent they can embed, configure, and trust — one that operates inside their applications with the security and control their organizations expect.”
What’s new: the document editing agent
- Agentic execution inside the viewer: A reasoning layer plans work, selects tools, and carries multi-step document tasks through to completion using the viewer’s high-fidelity rendering and document structure.
- Observe, verify, adapt: The agent executes actions in sequence, observes results, verifies outcomes, and adjusts as needed across extraction, form filling, annotation, and redaction workflows.
- Staged changes with approval: Edits are staged for review and applied permanently only after approval, keeping humans in control of the final output.
- Policy-governed autonomy: Organizations define what the agent knows (skills), how it operates (context), and what it can do (approval policies) – from fully autonomous actions to hard prohibitions.
- Not a single-shot feature: A tool-using system designed to execute document work reliably inside the application where users already operate.
- Built for document-heavy industries
AI Assistant is designed for industries where accuracy, compliance, and auditability are critical, including:
- Legal: Extract liability caps and indemnification clauses from a contract, then stage redactions of client names for attorney review before permanent removal.
- Financial services: Read a receipt and populate an expense report automatically. Extract invoice line items or redact bank account numbers before audit sharing.
- Construction: Render a blueprint, extract door dimensions from the second-floor plan, pull window schedules from tables, and highlight fire-rated material references.
- Healthcare: Extract patient demographics from intake forms, summarize lab report findings with citations to specific pages, or redact PHI before sharing with research teams.
- Government: Identify personal identifiers across a public records request, stage redactions for review, and apply permanent removal on confirmation.
Your business logic, embedded in the agent
Document automation only creates value when it behaves the way a business operates. This is especially true in regulated, document-heavy work where “almost right” is a risk. Nutrient AI Assistant is built so organizations can apply their rules to every workflow and get consistent outcomes across teams, documents, and applications.
For developers, this means they can deliver that consistency inside their application without building and maintaining a bespoke document agent stack. They embed the Viewer SDK and configure governance controls so higher-risk actions follow prescribed approval paths, while lower-risk steps can run automatically. The result is workflow automation that matches how the business works and still fits the organization’s compliance and review model as requirements evolve.
Ultimately, Nutrient AI Assistant helps companies turn document work from a bottleneck into a repeatable, policy-aligned process that improves speed, reduces risk, and increases operational consistency.
Source: PRNewswire