Squint uses AI to automatically transcribe your video, identify distinct steps, and structure them into a formatted procedure. You simply record yourself performing the task while narrating, and Squint handles the rest , no manual transcription or formatting required.

The transcript is generated automatically from your video’s audio using AI-powered speech recognition. There is no need to write or upload a separate transcript.

Squint supports languages across three tiers: Procedure Generation & Caption Generation (34 languages), App Interface for Mobile and Web (10 languages), and Mobile Procedure Translation (58 languages) for distribution to multilingual teams.

No. While video is the most common input, Squint can also generate procedures from existing scripts, PDFs, and other document formats. If you have legacy documentation, it can serve as a starting point.

Yes. Your company retains full ownership of all procedures created in its Squint workspace, including after your contract is terminated. You may export your data at any time during the term of your agreement.

Squint is built for real-world conditions. Imperfect recordings, occasional pauses, or off-script narration will still produce a usable first draft. You can review and edit any step after generation to refine the output.

Your team captures the video. Squint processes it from there. We provide best-practice guidance on recording to help maximize output quality on the first pass. Customers on the Premier Success Plan also receive hands-on support, including live training sessions and a 2-day onsite visit.

Squint includes tools to help identify and manage sensitive content. Please consult your Squint administrator or account representative for details on your organization’s specific data handling configuration.

Yes. Every step is fully editable after generation. You can reorder, rename, add, or delete steps, adjust formatting, and convert the procedure into a checklist format. Squint also includes a Find & Replace tool that lets you update terminology across all steps at once.

Yes. Procedures can be displayed as step-by-step instructions or as interactive checklists. Each step can include smart modules such as photo evidence upload, signature capture, numeric input, multiple choice questions, or pass/fail verifications.

Procedures can be updated directly within the Squint platform at any time. All changes are versioned so you can track what was updated and when. For organizations that require a formal review process, Squint supports Document Control, which routes drafts through designated approvers before changes go live.

Squint’s speech recognition is designed to handle a range of accents and background noise. For best results, recording in a reasonably quiet environment and speaking clearly will improve the quality of the first-pass output. For loud environments we recommend using Bluetooth lavalier (lapel) microphones.

Yes. Squint is fully accessible via web browser on desktop, no software installation required. Simply go to https://app.squint.ai

Unlike general-purpose AI tools, Squint is purpose-built for industrial and operational procedures. It understands the context of physical, hands-on tasks and connects directly to your operations through features like AR guidance and Assistant, all grounded in your organization’s actual content.

Yes. Existing PDFs and documents can be uploaded and used as source material for generating on-the-job guidance or refining procedures.

Yes. Previously recorded videos can be uploaded and processed through Squint’s AI just like new recordings.

A few simple practices make a meaningful difference: narrate each step clearly as you perform it, minimize background noise, use steady camera work, and record in manageable segments rather than one long take.

Accuracy is high and improves with video and audio quality. Yes, you can review each generated step alongside the corresponding video segment to verify and adjust as needed. The original video will be available on your device in your media library.

No, you do not need to record in a single take. Squint can process multiple video segments and handles pauses automatically. Please contact your Squint representative for current video length and file size limits.

One person can capture and narrate simultaneously. Hands-free recording is supported depending on your device setup, compatible wearables and smart glasses allow operators to record without holding a device. Squint also includes an AI Voiceover feature, so speaking during recording is entirely optional.

Yes. Squint works on both mobile devices and web browsers, so your team can access and perform procedures wherever they are, at a workstation or on the floor.

SOPs (Standard Operating Procedures) provide high-level process guidance, while work instructions offer granular, step-by-step direction for specific tasks. Squint supports both. You can author and perform either type depending on your operational needs.

Squint is valuable for both learning and ongoing use. New employees use it during onboarding to get up to speed quickly. Experienced operators use it to ensure compliance, reduce errors on infrequent tasks, and maintain consistency across the team.

Tasks are assigned units of work that connect an operator to a specific procedure. Admins or supervisors create a task, attach the relevant procedure, and assign it to an individual or team. Operators see their assigned tasks in a dedicated task list and know exactly what needs to be done.

Open the Squint app and go to Your Tasks. You can search by task name, reference number, or asset identifier. Tap the task to open it, then tap the attached procedure to begin.

Yes. Squint can integrate with external systems such as CMMS or ERP platforms to automatically generate tasks from work orders. Squint checks for new work orders on a regular interval and creates the corresponding tasks with the correct procedure attached, no manual steps needed.

Yes. Supervisors can create a task directly in Squint, attach the appropriate procedure, and assign it to a specific operator. This is useful for ad-hoc work that doesn’t originate from an external system.

Yes. Supervisors can reassign both the performer (the person doing the work) and the approver directly within the task in Squint. This is useful for shift changes, visiting team members, or when the originally assigned person is unavailable.

The task moves to a pending approval state. The assigned supervisor is notified that the task is ready for review. The supervisor reviews the completed procedure, including all inputs, photos, notes, and any flagged items, and either approves or reopens it.

This depends on how your organization configures the workflow. Tasks can be set to require supervisor sign-off before the record is considered complete, which is common for compliance-driven processes like inspections and audits.

The supervisor can reopen just the relevant section of the task. The operator sees it return to their task list with a note about what needs correcting. They fix only the flagged section and resubmit, no need to redo the entire procedure.

Yes. If you are set up as an approver for multiple sites, you can see pending approvals across all of them in your task list, as long as you have access through the appropriate Access Groups.

Yes. Go to Settings in the Squint app and enable notifications for new task assignments, completions, and approvals. This is the best way to know immediately when new work is assigned or when a task needs your attention.

Yes. When tasks are created from an external system, Squint automatically pulls in relevant details, such as asset information, work order numbers, and job details, and pre-populates them in the task. For most workflows, operators can open their task and get straight to work.

Completed and approved task records are stored in Squint and can be exported or automatically sent back to your external system of record. The exact export behavior depends on your integration setup, contact your Squint representative for details.

Not at the moment. Please contact your Squint representative for the latest information and detailed use case.

Squint uses a combination of visual recognition and location-based anchors to distinguish between similar equipment. Unique identifiers are applied to ensure the correct procedure is surfaced for the right machine.

AR anchors can be updated or recalibrated when equipment is moved or the workspace layout changes. This can be done without regenerating the full procedure.

AR guidance is most effective for complex, repetitive, or high-stakes tasks where step-by-step visual overlays reduce errors and improve speed, such as equipment maintenance, assembly, quality inspections, and troubleshooting.

AR is optional; Squint works as a full digital platform without it. For teams using AR, it is designed for frontline workers with minimal training. Inside an AR-enhanced procedure, simply point your device at equipment to surface the relevant procedure automatically.

No. Your operational data, procedures, and AR configurations are not shared with other Squint customers. Please refer to your service agreement for full data privacy details.

Area maps are stored for one year. Re-entering a mapped area resets the retention timer for an additional year, allowing frequently used locations to persist indefinitely. If an area remains inactive for a full year, the map data will be considered expired and the area will need to be recaptured.

Currently, the most practical solution is to use a MagSafe-compatible magnetic mount case, which lets you attach your device directly to the machine or equipment so operators can follow along without holding it. Squint is actively developing dedicated wearable device support for hands-free procedure execution.

Squint is designed to map focused sections of equipment rather than an entire machine or facility at once. The recommended approach is to map areas of approximately 5 x 5 feet (1.5 x 1.5 meters), concentrating on the most visually distinct and frequently accessed sections. For larger machines, simply create multiple mapped areas.

Unlike Microsoft Copilot which is a general-purpose productivity assistant built around Microsoft 365 applications, Squint Assistant is purpose-built for industrial operations. It is grounded in your organization’s procedures, equipment documentation, and operational knowledge base, delivering answers that are specific and directly relevant to your work.

ChatGPT is a general-purpose AI with no access to your company’s specific data. Squint Assistant is connected to your procedures and uploaded documents, so every answer is grounded in your organization’s actual content, not generic internet knowledge.

Assistant is not a procedure authoring tool, it won’t write an SOP from scratch. What it does is help operators and technicians quickly find the right procedure or document and know exactly what to do next. Think of it as an intelligent guide that points you to the right steps, at the right time, based on what you ask.

Yes. When you ask a question, Assistant can navigate you directly to the relevant step within a procedure, so you spend less time searching and more time doing.

Yes. Squint Assistant is connected to your procedures and uploaded documents. Answers are drawn directly from your organization’s knowledge base.

Yes. Assistant can reference uploaded schematics and technical diagrams to help answer questions.

Assistant can access your Squint procedures and any documents uploaded by your organization. Administrators can configure what content is available based on your operational needs.

Assistant currently supports PDF files for document uploads, including schematics and technical diagrams. Support for additional file types, including common Microsoft Office formats, is actively in development.

Squint Assistant is powered by leading large language models enhanced with Squint’s proprietary industrial knowledge layer. Specific model details are available upon request.

No, it needs connectivity to query your library in real time.

Squint is hosted on secure cloud infrastructure. Please contact your Squint representative for details on hosting regions and compliance certifications relevant to your organization.

Assistant automatically reflects changes when procedures or documents are updated in Squint. There is no manual retraining or republishing required.

No. Your data is used solely to power your organization’s Assistant experience. It is never made accessible to other customers.

Assistant automatically reflects the updated content in future responses. There is no delay or manual refresh needed.

When multiple relevant sources exist, Assistant surfaces the most applicable answer along with citations, allowing you to review the source material directly.

Users can be added manually via email invite, in bulk via CSV upload, or automatically provisioned through Verified Domains. With Verified Domains configured, any user who signs in with a company email is onboarded automatically, no manual setup required.

Squint uses role-based access control (RBAC). The core roles are Admin (full platform control), Content Author (create and edit procedures), and Operator (perform procedures). Permissions can be scoped to ensure each role only has access to what they need.

Yes. Access Groups allow you to segment users and content by department, location, or team. Operators only see the procedures and materials relevant to their role or group.

Yes. Operators can select their profile at login. Where PIN Codes are enabled, users can sign in quickly without needing an email address, ideal for shared devices on the floor.

Yes. Admins can upload a CSV to onboard large numbers of users at once, assigning roles and Access Groups as part of the import.

Content created by a departing team member remains in the platform. Admins can reassign ownership to ensure continuity.

Yes. Procedure Sets and Tasks allow you to group procedures and assign them to specific individuals or teams, giving each operator a curated list of what they need to perform.

Access Groups support a hierarchical structure, allowing you to organize by department, site, or team. Content is scoped so each group only sees what is relevant to them.

Yes. Squint supports SSO for organizations with verified domains, allowing users to sign in with their existing company credentials, no separate password required.

Content lives in the Library, organized into four tabs: Procedures, Procedure Sets, Areas, and Documents. Companies can also set up separate Workspaces for different business units or entities, entirely optional and up to each organization.

Yes, through user roles. Content Authors can create, edit, and publish procedures or submit drafts for approval. Admins review and approve content before it is published. You decide who on your team gets to create and approve.

Enterprise Collaboration allows corporate teams to create standardized procedures and distribute them across multiple Workspaces. Local teams access shared content without duplicating effort, and corporate teams retain visibility and control.

Yes. One approach is to create an Archive Access Group, reassign inactive procedures to it, and ensure no users are assigned to it. Procedures stay in the system and visible to admins but out of reach for operators and can be restored anytime if needed.

When a procedure is updated and republished, operators always see the latest version. Notifications can alert relevant users when content changes.

Yes. Procedure Sets let you bundle related procedures and assign them to users or teams, making it easy to structure onboarding paths, training tracks, or role-specific workflows.

The Analytics Dashboard gives admins visibility into procedure engagement, completion rates, and performance across users and Access Groups, helping identify underutilized content and areas for improvement.

Yes. Existing PDFs can be uploaded directly into Squint, where they become available to Assistant and operators as reference material.

Squint’s Find & Replace tool lets you search for and update specific terms across all steps in a procedure at once, useful for standardizing language or correcting transcription errors across an entire procedure.

Yes. Required Media allows admins to attach mandatory video assets to a procedure. Operators must watch the content in full before they can advance, ensuring everyone starts from the same baseline.

Squint’s integration capabilities are actively expanding. For organizations that want to connect custom systems, API Endpoints will be included with all Squint plans. Reach out to your Squint representative to discuss your specific environment.