Documents & Letter Writing

Letter writing is one of the most time-consuming tasks in private general practice. Every consultation that results in a referral, a report to a patient's NHS GP, a discharge summary, or a medico-legal letter creates a document that must be accurate, professionally formatted, and clinically complete. The difference between a platform that makes letter writing fast and one that makes it merely possible can mean an hour or more per clinician per day.

This comparison examines how Jump EHR and Semble handle clinical document creation, templates, auto-population, AI-assisted drafting, referral workflows, dictation, document storage, and correspondence distribution. Both platforms provide letter writing functionality, but they take fundamentally different approaches to how much of the work the software does for you - and those differences are among the largest in any feature area we've compared.

At a Glance: The Headline Differences

The table below summarises the key capability differences. Items where one platform has the clear edge are noted.

• Pre-built clinical letter templates - Jump: 5 included templates (Consultation Outcome, Referral Narrative, Results Reassuring, Results Further Review, Appointment Confirmation). Semble: No pre-built templates - create from scratch.
• Document type classification - Jump: SNOMED CT clinical document type codes. Semble: No clinical coding for documents.
• Merge field count - Jump: 90+ fields covering demographics, clinical data, practice, clinician, system. Semble: Basic demographics, signature image, practice logo, GP details.
• Clinical data in merge fields - Jump: Medications, allergies, problems, vitals, history, procedures, referrals, immunisations, consultation notes. Semble: Not available.
• Conditional logic in templates - Jump: {{#if}} syntax for adaptive content. Semble: Not available.
• AI letter drafting - Jump: Built-in, reads full clinical record, included in platform. Semble: No native AI - Heidi AI at £599/year per licence (separate product).
• AI template generation - Jump: Describe in plain language, AI generates template with merge fields. Semble: Not available.
• Rich text editor - Jump: Full formatting including tables, no restrictions. Semble: Basic formatting; tables only in templates, no page margin control.
• Google Docs / Word native editing - Jump: Full integration - create documents natively in Google Docs or Word. Semble: Word Add-In (cannot use auto-fields in text boxes).
• Referral letter workflow - Jump: Dedicated with recipient address book, priority levels (Emergency/2WW/Urgent/Routine), safety questions. Semble: Generic letter system - no dedicated referral workflow.
• Discharge summary workflow - Jump: Structured with medication reconciliation (NEW/STOPPED/CHANGED). Semble: Via Heidi AI only (£599/yr); otherwise manual.
• Headers and footers - Jump: Editable text with merge field support. Semble: Static uploaded images (Slim/Standard/Tall, 2MB cap).
• Document storage options - Jump: Native cloud + Google Drive + SharePoint. Semble: Single Documents tab with file upload.
• AI document filing (inbound) - Jump: AI extracts titles, suggests SNOMED codes, matches to patients. Semble: Not available.
• Voice transcription - Jump: Built-in with UK medical terminology vocabulary, included. Semble: OS-level dictation or Heidi AI (£599/yr) or unsupported Dragon.
• AI consultation structuring - Jump: Freeform notes to structured SNOMED-coded problems, allergies, tasks, document suggestions. Semble: Not available.
• Email integration - Jump: Gmail, Microsoft 365, Jump Mail - practice chooses provider. Semble: Secure link or email attachment; no email provider integration.
• SMS communication - Jump: Firetext integration. Semble: Not documented for letter distribution.
• Document export formats - Jump: DOCX, PDF, native Google Docs, SharePoint. Semble: PDF, secure link, print.
• NHS GP directory - Jump: Not documented. Semble: Built-in searchable directory (cannot auto-add GPs to records).
• Mail merge / batch letters - Neither platform offers this.
• Multiple recipients per letter - Jump: Not documented. Semble: Supported.
• Digital signature - Jump: Not applicable to letters. Semble: Image-based placeholder in letters; legally compliant for prescriptions.

Templates and Auto-Population: The Foundation of Every Letter

Starting from Scratch vs Starting with Structure

When a practice first sets up Semble, the Letters module is empty. Every template must be created from scratch - referral letters, GP correspondence, discharge summaries, patient reports, medico-legal letters. For a busy practice, building a complete template library is a significant upfront investment.

Jump ships with five pre-built clinical letter templates - Consultation Outcome, Referral Narrative, Results Reassuring, Results Further Review, and Appointment Confirmation - covering the most common document types in private general practice. These aren't generic stubs - they follow standard UK clinical correspondence conventions and include the appropriate merge fields for each document type. Practices can use them as-is, customise them, or create entirely new templates alongside them.

Both platforms support custom template creation. The difference is whether you start from zero or from a working baseline.

Merge Fields: Demographics vs the Full Clinical Picture

This is the single largest capability gap in the documents comparison. It determines how much of a letter the system writes for you versus how much you type manually.

Jump's template system supports over 90 merge fields spanning five categories. Patient demographic fields include title, full name, date of birth, age, NHS number, contact details, and individual address components. Clinical fields - the transformative category - pull in active medications with dosages, complete medication history, current supplements, all supplements, the last prescription record, formatted prescription tables, allergies with criticality ratings, active problems, full problem history, recent vitals and observations, social history, family history, recent procedures, active referrals, immunisation status, and consultation notes (last note, last 3, last 5, or all). Practice and clinician fields cover name, address, phone, email, and registration number. System fields include current date and time.

Semble's placeholder system covers basic demographic fields (name, DOB, address), a signature image, practice logo, and GP details. Clinical data - medications, allergies, problems, vitals, history - cannot be auto-populated into letters. Every clinical detail must be typed or copied manually.

To put this concretely: when a Jump clinician creates a referral letter using a template, the letter automatically contains the patient's current medication list, known allergies, active problem list, and relevant consultation history. A Semble clinician creating the same referral must type or copy-paste every clinical detail from the patient record into the letter. For a practice writing 20 referral letters a week, the cumulative time difference is substantial.

Conditional Logic: Templates That Adapt

Jump's merge fields support conditional logic using {{#if}} syntax. A template can include a medication section that only appears if the patient has active medications, or an allergy warning block that renders only when allergies are documented. This prevents letters from containing empty sections or misleading "No data" placeholders when information genuinely isn't recorded.

Semble's questionnaire builder includes conditional logic for patient-facing forms, but its letter template system does not support conditional content. Template sections are static - they appear regardless of whether the underlying data exists.

Template Management and Collaboration

Jump offers two template storage approaches: structured content stored in Jump's database, or native Google Docs files stored in the practice's own Google Drive. The Google Docs approach means templates can be collaboratively edited by multiple team members with full version history - a practice manager drafts a template, the lead clinician reviews and edits, and the final version is used across the practice. Every template can be assigned a SNOMED CT document type code for clinical classification.

Semble stores templates in its Settings area with a built-in editor. The Microsoft Word Add-In allows creating templates in Word and uploading them to Semble, but with the limitation that auto-fields cannot be inserted into Word text boxes (only Word frames). Collaborative editing and version history are not native features of Semble's template system.

AI-Powered Letter Writing: Built-In vs Bolt-On

This is where the two platforms diverge most dramatically.

Jump's Integrated AI

Jump includes a built-in AI writing assistant that drafts clinical letters using the patient's full clinical record. When a clinician creates a new letter, they can choose to have the AI generate a first draft based on the patient's active and historical problems, current and past medications with dosages, allergies and their criticality, recent vitals, social and family history, recent procedures, active referrals, immunisation status, recent consultation notes, the selected document type, and the intended recipient.

Because the AI reads directly from the structured clinical record, the draft is contextually accurate from the start. There is no copy-pasting consultation transcripts into a separate widget, no switching between applications, and no re-entering clinical information that already exists in the system.

The AI understands document structure. A referral letter draft follows proper clinical format: opening statement with a clear request, presenting complaint, relevant history, examination findings, investigations, current management, specific request, and urgency indication. A discharge summary is structured with admission details, diagnosis, clinical summary, medication changes clearly marked as NEW, STOPPED, or CHANGED, allergies, outstanding results, follow-up plan, and patient information provided. All AI-drafted content adheres to NHS standards: UK English spelling, professional tone, patient safety emphasis, standard abbreviations (PMH, O/E, Ix, Rx), pertinent negatives, and quantified findings.

If the AI draft isn't right, it's just a starting point - the clinician edits freely in the rich text editor. If the AI isn't wanted, the system falls back to a blank document or a standard template. Jump's AI features use OpenAI models (GPT-4o-mini for template generation and letter drafting, Whisper for voice transcription) - worth noting for practices with clinical data governance requirements around AI providers. Every AI interaction is audit-logged for clinical governance.

Jump's AI Template Generation

Beyond letter drafting, Jump uses AI to help practices create templates. Rather than building templates from scratch, staff describe what they need in plain language - "a referral letter template for dermatology with medication list and allergy section" - and the AI generates a complete template with appropriate merge fields, clinical structure, and formatting, ready to review and customise.

Semble's Approach: Heidi AI as a Separate Product

Semble has no native AI letter writing capability. Practices wanting AI assistance must purchase a separate Heidi AI subscription at £599 per year per licence (after a 30-day free trial). Heidi is an ambient AI scribing tool that transcribes consultations and generates structured notes from 25+ template types, pushing content into Semble's letter editor.

The integration works - Heidi can generate referral letters and discharge summaries from consultation transcriptions, and the content pushes into Semble letters. But the approach has inherent limitations compared to Jump's integrated model.

Heidi reads from the consultation transcript, not the structured clinical record. It knows what was discussed in this consultation but doesn't have access to the patient's full medication history, complete problem list, historical allergies, or previous consultation notes unless those happen to be mentioned in the current session. Jump's AI reads the entire structured record.

Heidi is a separate application. Clinicians must run Heidi alongside Semble, manage a separate subscription, and work across two interfaces. Jump's AI is embedded in the letter creation workflow - one system, one interface, one click.

Heidi does not generate templates. It generates letter content from transcriptions, but practices still need to build their own template library manually. Jump's AI generates both letters and templates.

Cost matters. For a five-clinician practice, Heidi AI adds £2,995 per year on top of the Semble subscription. Jump's AI is included in the platform price.

Referral Letters and Discharge Summaries

Referral Letters: Dedicated Workflow vs General Purpose

Jump has a dedicated referral letter workflow with recipient selection from an address book, priority level classification (Emergency, 2-Week Wait, Urgent, Routine), safety question fields, and specialist-appropriate formatting. Referral-specific merge fields include recipient name, title, specialty, organisation, and address. When AI consultation structuring is active, Jump can automatically suggest that a referral is needed based on the consultation content - including the suggested recipient specialty and urgency level.

Semble has no dedicated referral workflow. There is no "Create Referral" button, no priority level classification, and no structured routing to specialists. Referrals are handled through the general letter system using custom templates. Without Heidi AI, referral letters must be written entirely from scratch.

The practical difference: in Jump, a clinician finishes a consultation, clicks "Create Referral," selects the recipient from an address book, chooses urgency, and the system generates a clinically complete referral letter pre-populated with the patient's relevant history. In Semble, the clinician opens the general Letters module, selects a custom referral template (if one has been created), and manually populates the clinical details.

Discharge Summaries

Jump's discharge summaries follow a structured format with all required clinical sections and - critically - medication reconciliation clearly marked as NEW, STOPPED, or CHANGED. This is a patient safety feature: the receiving clinician can immediately see what changed during the episode of care without cross-referencing multiple documents.

Semble can generate discharge summaries via Heidi AI (at additional cost), but has no native structured discharge summary workflow. Without Heidi, discharge summaries are created as general letters.

The Writing Experience: Editors and Formatting

Both platforms provide rich text editors for letter writing. Jump's built-in editor supports bold, italic, strikethrough, font family and colour selection, text alignment, heading levels, bullet and numbered lists, and block quotes. For documents requiring tables, hyperlinks, or more advanced formatting, Jump's Google Docs and Microsoft Word integrations provide full-featured editing within the platform - documents are real Google Docs or Word files with the complete formatting toolset.

Semble's editor offers similar basic formatting (bold, italic, underline, tables, hyperlinks, paragraph spacing, image insertion) but with documented limitations: tables can only be inserted in templates, not in individual letters, and page margin customisation is not available.

For practices that prefer working in familiar tools, Jump offers a significant advantage: documents can be created natively in Google Docs or Microsoft Word with the full formatting capabilities of those platforms. This isn't a file export - it's native integration where the document is a real Google Doc or Word document, editable with all the features those platforms offer, and connected back to the patient record. Semble offers a Microsoft Word Add-In, but it cannot insert auto-fields into Word text boxes (only Word frames), limiting its utility.

Branding, Headers, and Footers

Professional appearance matters in private practice correspondence. Both platforms support organisation-level branding, but the implementations differ.

Jump's headers and footers are editable text with merge field support - practices can include dynamic content like the clinician's name, practice address, and registration number that updates automatically. Font family, font colour, logo placement, and custom content are all configurable at the organisation level and embedded automatically into every document. Document watermarking is available for draft or confidential markings.

Semble requires headers and footers to be uploaded as static images in Slim, Standard, or Tall sizes with a 2MB file cap. This means any change to header content (a new phone number, an updated logo, a changed address) requires creating and uploading a new image file rather than editing text. Merge fields cannot be used in headers or footers.

Document Storage and Filing

Storage Options

Jump offers three document storage approaches to suit different practice setups. Native cloud storage provides secure document management within Jump, supporting any file type up to 50MB with organisation-by-patient folder structures. Google Drive integration creates a dedicated "Jump EHR Patient Records" folder structure with patient-specific subfolders - documents are native Google Docs with collaborative editing and version history. Microsoft SharePoint integration provides the same structured storage for practices using Microsoft 365.

Semble offers a single Documents tab within each patient record with file upload, drag-and-drop, folder organisation, and search/filter by type and date. This covers the basics but does not provide the deep integration with external document management platforms that Jump offers.

The practical implication: practices already using Google Workspace or Microsoft 365 can work within the tools they know, with patient documents living alongside other practice files in a familiar environment. Practices that prefer everything in one system can use Jump's native storage.

AI-Powered Inbound Document Filing

Jump includes an AI-powered document filing system with no equivalent in Semble. When documents are uploaded (referral letters from specialists, lab results, clinical correspondence from hospitals), AI analyses the content - extracting document titles, suggesting SNOMED CT codes, and matching documents to patients automatically. Documents move through a managed filing workflow (Unmatched, Matched, Filed) with human oversight at each stage.

For practices that receive dozens of inbound letters and results per day, this transforms document filing from a manual sorting exercise into a review-and-confirm workflow. Semble's inbound document handling is manual upload and organisation only.

Dictation and Voice Transcription

Jump includes native voice transcription purpose-built for UK clinical use, with a UK medical terminology vocabulary covering NHS clinical terms, standard abbreviations (PMH, DH, SH, FH, O/E), common conditions, medications, examination findings, and red flag symptoms. The transcription uses OpenAI's Whisper speech recognition with a clinical prompt layer that ensures accurate capture of medical terminology. Recording happens directly in the browser with echo cancellation and noise suppression. Transcribed text can be inserted into consultation notes and used as the basis for AI-powered letter drafting.

Semble's dictation options are fragmented: OS-level browser dictation (Mac/Windows built-in speech recognition with no medical vocabulary), Heidi AI at £599 per year per licence for ambient AI scribing, or theoretically Dragon Medical (not officially supported or integrated). The OS-level option is free but unreliable for clinical terminology. Heidi is capable but adds cost and complexity. Dragon requires separate software and has no native integration.

Jump's transcription is integrated, medically optimised for UK clinical language, and included in the platform price.

AI Consultation Structuring

Beyond dictation, Jump's AI consultation structuring takes freeform or voice-transcribed notes and transforms them into structured clinical data - extracting multiple problems with SNOMED CT coding, identifying allergies, flagging actionable tasks (referrals, investigations, follow-ups), and suggesting documents that need to be written. This last capability is a distinctive workflow advantage: the AI identifies when the consultation implies correspondence is needed, suggests the document type (referral letter, discharge summary, results letter), and the clinician can launch directly into the letter wizard with the clinical context pre-loaded. A clinician dictates their consultation notes, and the system not only transcribes and structures the clinical content but actively drives the follow-up correspondence workflow.

This goes well beyond what Heidi AI offers through its Semble integration. Heidi transcribes and generates notes; Jump's system transcribes, structures, codes, identifies actions, and triggers document creation workflows.

Correspondence Distribution

Jump integrates with Gmail, Microsoft 365, and its own Jump Mail service for letter distribution - practices choose their preferred email provider, and letters are sent from the practice's own email domain. SMS communication is available via Firetext integration. A unified communication method selector lets staff switch between email and SMS depending on the recipient and context. Documents can be exported as DOCX or PDF with full branding.

Semble restricts letter sharing to secure links, email attachments (PDF), or print. There is no native email provider integration - letters are sent from Semble's system rather than the practice's own email domain. Recipients must be pre-stored contacts (free-typing email addresses is no longer allowed as a security measure). There is no NHSmail or MESH integration. Neither platform integrates with NHSmail or MESH.

Semble does have a built-in NHS GP Directory searchable by postcode, practice name, phone, email, or website - useful for looking up referring GP details. However, GPs found in the directory cannot be directly added to patient records; the information must be manually copied. Jump does not document an equivalent built-in NHS GP directory.

Where Semble Has the Edge

• NHS GP Directory. Semble has a built-in searchable directory of NHS GP practices, useful for looking up contact details when addressing correspondence. Despite the limitation that GPs cannot be auto-added to patient records, having an integrated directory is more convenient than searching externally. Jump does not document an equivalent feature.
• Multiple recipients per letter. Semble supports sending a letter to multiple recipients simultaneously. Jump's multi-recipient capability is not documented.
• Semble's letters case study. Semble has published customer evidence showing real time savings - Tracy Sutton of TAS Secretarial Services reported saving four to five hours per week, and Schoen Clinic reported reducing letter writing time from three hours to 45 minutes (likely with Heidi AI). These are credible, specific claims that demonstrate Semble's letters module works well for practices that have invested time in building their template library.

The Bottom Line

Documents and letter writing is the comparison area where the gap between Jump and Semble is widest. The differences are not marginal - they represent fundamentally different approaches to how much clinical letter writing the software automates versus how much falls to the clinician.

Jump is the stronger choice for practices that need: AI-powered letter drafting from the clinical record (without a separate subscription), clinical data auto-population in templates (medications, allergies, problems, vitals, consultation notes), dedicated referral and discharge summary workflows with structured formatting, SNOMED CT document classification, native Google Docs or SharePoint integration for collaborative editing, AI-powered inbound document filing, integrated voice transcription with medical terminology, or flexible email integration through Gmail or Microsoft 365.

Semble is the stronger choice for practices that: already have Heidi AI subscriptions and are satisfied with the bolt-on AI approach, have invested heavily in building their own Semble template library, need the built-in NHS GP directory for addressing correspondence, or prefer a simpler document system where the basics (create letter, add patient details, send) work without the additional capabilities.

For clinicians who write multiple clinical letters per day - and in private general practice, most do - the combination of 90+ merge fields with clinical data auto-population, integrated AI drafting, and dedicated referral workflows means Jump can reduce letter writing from a 10-15 minute manual task to a 2-3 minute review-and-send workflow. That time saving, multiplied across every consultation that generates correspondence, is the single most impactful productivity difference between the two platforms.

Methodology

This comparison was compiled in February 2026 using Semble's public help centre documentation (help.semble.io), product pages (semble.io), verified user reviews from Capterra, GetApp, and Software Advice, Semble's published case studies, and direct inspection of Jump EHR's codebase and product functionality. Where a feature is described as "not documented" for either platform, it means we could not find public evidence of the feature - it may exist but not be publicly visible. We encourage readers to verify specific capabilities during a product demonstration.