Artificial Intelligence

Lojycal Intelligence that reasons over your whole IT stack.

Eleven AI surfaces, one reasoning fabric. Lojyc AI, Script Generator, The Investigator, AI Analysis Engine, AI Subscription Analyzer, The Builder, The Negotiator, AI Policy Generator, Process Documentation, Savings Bot and AI Scan PDF — every one cites the in-app workspace surface it powers.

One reasoning surface. One audited action trail. One source of truth.

Why "AI features" usually disappoint

Most platforms bolt a chat sidebar onto disconnected tables. The model can summarise a row, but it can't see the contract that explains the renewal, the IdP signal that explains the device, or the vendor benchmark that explains the price. So its answers stay shallow and its "actions" never leave the textbox.

Lojycal AI is wired into the same governed ledger as the rest of the product — identity, posture, finance, contracts, audit. It reasons across them, proposes actions, and writes every step to the WORM log. You get decisions, not paragraphs.

Where AI shows up in Lojycal

Ten primary surfaces, one reasoning fabric. Each capability cites the in-app surface it powers — open the workspace and you'll find it there.

Lojyc AI

Lives in: the floating Sparkles button, every page

Your in-platform copilot. Tier-aware — it knows what your workspace can and can't do, gives exact click paths, drafts the next step, and recommends the upgrade only when a locked module would genuinely help.

Links to: every module, billing tier, navigation.

Script Generator

Lives in: AI Patch Lab → Script generator

Generates production-ready PowerShell and Bash patch/automation scripts with a built-in validator (risky cmdlets, CRLF, non-ASCII, #Requires, bracket/alias checks), required-params block, dry-run, then Run script straight to your RMM.

Links to: Patch Library, Automation Library, Automation Queue, Endpoint Governance.

The Investigator

Lives in: AI Threat Lab → The Investigator

Plain-language threat analyst. Correlates signals across SaaS, Endpoint and Identity, cites the governing policy from Policy Lab, and exports the answer as Copy / PDF / RTF for the incident pack.

Links to: Shadow IT Lab, Endpoint Governance, IdP signals, AI Policy Generator.

AI Analysis Engine

Lives in: Contract Lab → AI Analysis Engine

Drop a contract PDF and the engine extracts the renewal date, TCV and indemnification clauses, then surfaces HIGH / MED / LOW risk flags for unfavourable terms — before renewal lands on the CFO.

Links to: Filing Cabinet, Financials renewals, Procurement Lab.

AI Scan PDF

Lives in: Contract Lab → Filing Cabinet → AI Scan PDF (also AI Scan Invoice in the Invoice Database)

Drop any contract or invoice PDF and the scanner reads the document end-to-end — extracting counterparty, renewal date, total contract value, currency, indemnity and liability clauses, and line-item amounts — then auto-files the original PDF into the Filing Cabinet (or Invoice Database) with the parsed fields pre-populated. Every extraction is written to the WORM audit_log so finance and audit see the same source of truth.

Links to: Filing Cabinet, Invoice Database, AI Analysis Engine, Procurement Lab, Financials renewals.

The Builder

Lives in: Contract Lab → The Builder

Prompt a contract, watch it draft live on the right. DPAs, MSAs, 60-day renewal-notice windows, liability caps tightened to 1× annual fees — one click to add to the Contract Cabinet or download.

Links to: Contract Cabinet, Procurement Lab, vendor onboarding.

The Negotiator

Lives in: Contract Lab → The Negotiator

AI counter-offer drafter. Compares your vendor's pricing and clauses against the global benchmark median and writes the redline + email you'd send back.

Links to: Financials benchmarks, Procurement Lab, Savings Bot.

AI Policy Generator

Lives in: Policy Lab → AI Policy Generator

Pick a Source Lab (Endpoint, Domain, Shadow IT…) and a framework (ISO 27001, SOC 2, NIS2, DORA) — the AI drafts a formal company policy from the lab's currently active settings, ready for the audit pack.

Links to: every Lab's live config, Audit Evidence Center, Trust Center.

Process Documentation

Lives in: Policy Lab → Process Documentation (Enterprise)

Generates SOPs and runbooks from your live workspace state — onboarding, offboarding, incident response, patch windows — so your tribal knowledge becomes the document the next admin actually reads.

Links to: Employee Workflows, Audit Evidence Center, Incident Reports.

Savings Bot

Lives in: Financials → Savings Bot

Your savings advisor. Answers "Where can we save the most this quarter?", "Which cost center is overspending?" and "Quick wins under €5k effort?" using live SaaS, CapEx and benchmark data.

Links to: SaaS licence inventory, Contract Lab renewals, global pricing benchmarks, Procurement Lab.

AI Subscription Analyzer

Lives in: Contract Lab → AI Subscription Analyzer

Drop a raw CSV / XLSX billing export and the analyzer returns a 3-colour verdict (🔴 Critical Bleed / 🟠 Operational Drift / 🟢 Stable Tier) in seconds — flagging zombie seats, cluster overlaps and seat-vs-SSO mismatch. Raw rows are purged from memory the moment the verdict renders.

Links to: The Negotiator, Offboarding Guard, Procurement Lab, ERP ledger.

Glossary

Plain-language definitions for every Artificial Intelligence surface in Lojycal, and how each one connects to the rest of the platform.

Lojyc AI
The in-platform copilot (the floating Sparkles button). Tier-aware navigation and how-to guide — knows what your workspace can do, gives exact click paths, and only recommends upgrades when a locked module would genuinely help.
Connects to: Reads your billing tier and module catalogue; deep-links into every Lab and admin surface.
Script Generator
AI author for patch and automation scripts. Generates production-ready PowerShell and Bash with a built-in validator (risky cmdlets, CRLF, encoding, #Requires, bracket/alias checks), required-params block and dry-run before anything runs on a fleet.
Connects to: Lives in AI Patch Lab; pushes approved scripts to the Patch Library, Automation Library, Automation Queue and out to the RMM.
The Investigator
Plain-language threat analyst. Ask 'what happened on this device last night?' and it correlates signals across SaaS, Endpoint and Identity, cites the governing policy, and exports the answer as Copy / PDF / RTF for the incident pack.
Connects to: Pulls from Shadow IT Lab, Endpoint Governance and IdP signals; cites Policy Lab; writes findings into Incident Reports.
AI Analysis Engine
Contract reader. Drop a contract PDF and the engine extracts renewal date, total contract value and indemnification clauses, then flags HIGH / MED / LOW risk on unfavourable terms before renewal lands on the CFO.
Connects to: Lives in Contract Lab; feeds renewals into Financials and procurement decisions into Procurement Lab.
AI Scan PDF
Document reader for contracts and invoices. Drop a PDF into the Filing Cabinet (AI Scan PDF) or Invoice Database (AI Scan Invoice) and the scanner extracts counterparty, renewal date, total contract value, currency, indemnity and liability clauses, and line-item amounts, then auto-files the original PDF with the parsed fields pre-populated on the record. Every extraction is written to the WORM audit_log with actor and reasoning hash.
Connects to: Lives in Contract Lab; feeds parsed contracts into the AI Analysis Engine, renewals into Financials, and procurement decisions into Procurement Lab; invoice scans land in the Invoice Database for finance reconciliation.
IT Contract Builder (The Builder)
AI drafter for IT contracts. Prompt a DPA, MSA or vendor agreement and watch it draft live — 60-day renewal-notice windows, liability caps tightened to 1× annual fees, GDPR sub-processor language — one click to add to the Contract Cabinet.
Connects to: Lives in Contract Lab; outputs land in the Contract Cabinet, Procurement Lab and vendor onboarding flows.
The Negotiator
AI counter-offer drafter. Compares your vendor's pricing and clauses against the global benchmark median and writes the redline plus the email you'd send back to the account manager.
Connects to: Reads Financials benchmarks and Savings Bot signals; pushes redlines into Contract Lab and Procurement Lab.
AI Policy Generator
Drafts formal company policy from your live workspace state. Pick a Source Lab (Endpoint, Domain, Shadow IT…) and a framework (ISO 27001, SOC 2, NIS2, DORA) and the AI writes the policy from the lab's actual current settings — not a generic template.
Connects to: Lives in Policy Lab; reads every Lab's live config; outputs feed the Audit Evidence Center and Trust Center.
Process Documentation
Generates SOPs and runbooks from your live workspace — onboarding, offboarding, incident response, patch windows — so tribal knowledge becomes the document the next admin actually reads. Enterprise tier.
Connects to: Lives in Policy Lab; reads Employee Workflows and Audit Evidence Center; output indexed alongside Incident Reports.
Savings Bot
Your live savings advisor. Answers 'where can we save the most this quarter?', 'which cost centre is overspending?' and 'quick wins under €5k of effort?' using current SaaS licences, CapEx, contracts and global pricing benchmarks.
Connects to: Lives in Financials; reads the SaaS licence inventory, Contract Lab renewals, global pricing benchmarks; hands work to Procurement Lab and The Negotiator.
AI Subscription Analyzer
Auditor-grade ephemeral billing-data analyzer. Drop a raw CSV/XLSX export and the engine parses vendor strings, currency-aware amounts and department codes entirely in browser memory — fuzzy header detection plus Levenshtein vendor matching means mixed headers, empty rows and unusual ledger formats still produce a 3-colour verdict (🔴 Critical Bleed / 🟠 Operational Drift / 🟢 Stable Tier). Raw rows are purged from memory the moment the verdict is rendered; a visible 'Data purged' pill confirms it. The frosted gate routes Elite Tier requests through a fully validated qualification form (work-email only, honeypot, rate-limit) straight to admin@lojycal.space. Inside the authenticated workspace each flagged row exposes a SCIM de-provision dry-run preview before any write, and confirmed actions are recorded in a chronological ERP remediation timeline (Phase 1 / 2 / 3).
Connects to: Saves money by: reclaiming zombie seats (typical £40–£180/seat/mo), killing duplicate cluster spend (HubSpot+Intercom, Slack+Teams, Figma+Sketch), flagging seat-vs-SSO mismatches that signal over-licensed contracts, and pushing every confirmed action into the ERP ledger so finance sees the recovery. Lives in Contract Lab → AI Subscription Analyzer; reads identity (Okta, Entra, GWS, M365) and MDM (Jamf, Intune, Kandji) telemetry, the contract Filing Cabinet and SaaS licences; fires SCIM revokes via Offboarding Guard and pushes renewal dates into Procurement Lab.
SQL Lab
Natural-language to safe read-only SQL against your own tenant. Ask "which laptops are off-warranty in Q4?" and the AI writes the query, previews the rows, and lets you save it to the Report Library for re-runs — all tenant-scoped, never with a god-mode key.
Connects to: Lives in Trust Center → SQL Lab; feeds Report Library, Audit Evidence Center, and every tenant-scoped operational table.

How AI plugs into the rest of the platform

Signals in → reasoning → actions out. Nothing happens in a side-channel; every input and every output lives in the same governed ledger your auditors already trust.

  1. Step 1
    Signals in

    Identity, MDM, SaaS, contracts, spend, audit_log and global benchmarks — read with per-tenant RLS, never with a god-mode key.

  2. Step 2
    Reasoning

    Each AI surface runs through the Lovable AI Gateway. Prompts, tools and context are scoped to your org; no customer data trains the underlying models.

  3. Step 3
    Actions out

    Approvals, automations, scripts, vendor pushes, policies and evidence-pack signatures — every AI-suggested action is logged, attributable and reversible.

Better decisions, sharper finance

AI is only useful when it changes the number. Lojycal AI rightsises licences, benchmarks vendor pricing, forecasts burn and prefills incident reports — so the CFO sees less surprise and the IT lead sees less manual work.

  • Licence rightsizing flags inactive seats with last-sign-in evidence.
  • Vendor pricing is scored vs. the global benchmark median in real time.
  • Burn forecasts react live to hires, renewals and patch waves.
  • Renewal alerts arrive 60+ days early with the clause that matters quoted.
  • Procurement requests get a green/orange/red verdict before approval.
  • Incident reports prefill from the WORM archive — minutes, not days.

Governance & safety, on by default

AI without governance is a liability. Lojycal AI runs inside the same RLS, MFA and audit rules as the rest of the platform — not in a parallel universe.

  • Every AI action is written to the WORM audit_log with actor + reasoning hash.
  • No customer data is used to train foundation models.
  • EU residency by default; secret columns stay column-encrypted.
  • Tier-gated: deeper reasoning (Investigator, Burn forecast) on Growth and above.
  • Privileged AI tools (kill-switch, evidence sign) require AAL2 + second-admin review.
  • Prompts and tool calls are observable per org — explainable, not magic.

AI should make Lojycal sense.

Not a chat sidebar. Not a shiny demo. A reasoning surface wired into identity, posture, contracts and finance — with every action signed, attributable and reversible.