The same private agent, shaped to how each team works.
AssistantGeneral ships with built-in personas and a policy engine, so the agent that helps Legal review a contract is the same install that helps Finance close the books — just configured differently, and never sharing data between them unless you say so.
Legal & compliance
Contract review without a document leaving the firm.
The challenge. Reviewing NDAs, vendor terms and lease renewals against firm standards is slow, repetitive, and the documents are too sensitive to paste into a public chatbot.
With AssistantGeneral. Point the Lawyer persona at a contract folder. It flags deviations from your standard clauses, surfaces renewal and termination dates across hundreds of files, and drafts a redline — citing the exact page and clause it used. Nothing leaves the firm's machines.
Built-in Lawyer persona, tuned for contract language
RAG over your own template library and precedent
Generates real, audited .docx redlines
Healthcare & life sciences
Literature review that cites real, current sources.
The challenge. Clinicians and researchers need fast, trustworthy answers grounded in current literature — and patient data can never touch a third-party cloud.
With AssistantGeneral. The Doctor persona queries PubMed, Europe PMC and ClinicalTrials.gov directly, cross-references your internal case notes through the local knowledge base, and flags retracted papers automatically. Every claim is sourced.
Direct connectors to PubMed, Europe PMC, ClinicalTrials.gov
Built-in retraction screening
Patient records never indexed off-premises
Financial services
Close-of-month reporting in minutes, not days.
The challenge. Analysts spend days reconciling spreadsheets from a dozen sources before a single report goes to the board.
With AssistantGeneral. Connect your read-only database and reporting folders. AssistantGeneral pulls the numbers, builds the workbook with the formulas already computed, audits its own figures against the source, and hands you a finished .xlsx — not a description of one.
Self-audited Excel generation with computed formulas
Full inference trace for audit trails
Manufacturing & engineering
One assistant across plant floor and head office.
The challenge. Technical documentation, maintenance logs and supplier contracts live in different systems no single tool reads end to end.
With AssistantGeneral. Connectors sync SharePoint, shared drives and your ticketing system into one knowledge base. The Mechanic persona answers maintenance questions on the floor; the same install handles supplier contract review in procurement — governed by separate policies.
20+ connectors unify scattered documentation
Per-group policy keeps plant-floor and office access separate
Runs fully on-prem for air-gapped facilities
Public sector & education
A capable assistant that never sends citizen data anywhere.
The challenge. Public bodies want the productivity of modern AI but cannot accept a vendor cloud touching constituent or student records.
With AssistantGeneral. The Government edition packages local-first inference, full audit, and a sovereign deployment model — so staff get a real agent without a single record leaving the building, satisfying procurement and data-residency requirements outright.
Sovereign, air-gapped deployment
Zero runtime cloud dependency
Centrally governed policy across the whole fleet
Defense & critical infrastructure
High-security work, fully offline.
The challenge. Highly regulated and network-restricted environments need AI capability with zero tolerance for network egress, by policy and often by law.
With AssistantGeneral. Run entirely air-gapped after installation. Every model, tool and connector operates from local infrastructure, with PIN lock, encryption at rest and an append-only audit log satisfying the strictest review.
No runtime network dependency, verifiable by isolation
AES-256-GCM encryption at rest
Append-only audit trail for every action
Messaging & Communication Ops
A Slack, Teams, or WhatsApp bot that reads before it acts — on your own hardware.
The challenge. Teams want an intelligent bot that classifies message intent and routes accordingly — open a ticket, notify on-call, auto-reply — but routing every Slack message through a third-party cloud is a security non-starter.
With AssistantGeneral. Configure a messaging bridge in minutes. Each incoming message is scanned by the local model — urgency, topic, sentiment, who should handle it — then the right workflow fires automatically: create a support ticket, page the on-call engineer, send an auto-reply, or all three in sequence. Everything runs on your hardware with a full per-message audit trail.
Inbound message → local AI classification → workflow fork
Auto-reply rules plus full agentic workflow routing
Same RBAC policy and audit log on every channel
Email Triage & Automation
Your inbox, triaged and handled by a local agent that never sends your email to a cloud.
The challenge. Parsing, categorising and acting on high-volume inboxes — vendor emails, executive correspondence, support requests — takes hours of manual work and the content is far too sensitive to pipe through a public AI service.
With AssistantGeneral. An event trigger watches a Gmail or IMAP inbox on a configurable poll interval. When a matching message arrives, the workflow summarises it, invokes the local LLM with a persona and any relevant knowledge base context, drafts a reply or action, and sends it — the full run logged with timing, input files, and output. The 'Boss Handler' pattern shown in the visual builder is a live example of this.
Gmail / IMAP event trigger with configurable poll interval and match rules
Full run log: input files, LLM response, duration, trigger reference
Zero email content leaves the building at any stage
Scheduled Reporting & Briefings
Daily briefings that pull from every source and land on your desk — composed by the local agent.
The challenge. Executives and team leads spend the first hour of every day aggregating updates from dashboards, shared drives, and emails into a coherent picture that could have been done automatically overnight.
With AssistantGeneral. A scheduled workflow runs on your cadence — daily, hourly, or after a trigger event. It pulls from connected data sources (databases, cloud storage, email, web), runs the synthesis through the local LLM, and produces a formatted digest: what changed, what needs attention, what can wait. No cloud API key, no per-token cost creep, no data leaving the building.
Schedule on any interval, cron expression, or event trigger
Pulls from databases, cloud storage, email connectors, and web sources
Chained sub-workflows: fetch → summarise → format → deliver
Fully editable digest template via the visual node editor
Don't see your team? It's still the same agent.
Personas, policy and connectors are configurable — most deployments take an afternoon to shape to a new department.
Talk to the team that actually builds the software.
Pilot deployments, volume licensing, product demos, security questionnaires — all handled by engineers and product leads, not a routing layer. We respond within one business day.
Schedule a discovery call
Half-hour walkthrough with someone who built the product — no sales script.
Run a pilot deployment
Full-feature evaluation with guided install, configured for your environment.
Email us directly
sales@royalsoftworks.com — we respond within one business day.