Ajelix

Ajelix is positioned as a persistent agentic AI workspace for business rather than only an Excel helper. It supports files/documents, persistent context, reusable asset and prompt libraries, dashboards, apps, presentations, spreadsheet work, data cleansing and recurring workflows. The main site states files up to 1GB and surfaces Google Workspace integrations. ¹

The Enterprise platform, generally available September 8, 2026, adds custom agents, visual workflows, document-grounded RAG and governance. First-party materials describe RBAC, audit trails, policy/guardrail enforcement, human escalation for high-stakes decisions, and Ajelix-hosted EU, self-hosted, or AWS/Azure/Google Cloud deployment. These are first-party claims; AiToolMap did not deploy or test the product. ^{2 3}

Persistent files/context combine with generated assets and prompt reuse; Enterprise adds schedule/event/webhook triggers, branching, loops, parallel execution and arbitrary data-source connections. Google Sheets, Docs, Slides and Forms integrations are surfaced on the site. ^{1 2}

The live pricing page lists 10 free requests/month. Monthly paid prices are Lite $39/user, Pro $99/user and Max $199/user; annual billing displays $32, $82 and $165/user/month. Enterprise is custom. Older G2/Capterra/TrustRadius pricing is stale and is not used as current price authority. ^{4 5 6 7}

The fixed 50-source panel produced 4 USE_WITH_LIMITS and 46 NO_EVIDENCE dispositions. Trustpilot has a small eight-review exact-company/product corpus with recent 2026 feedback, but much detailed feedback concerns spreadsheet/VBA/BI functionality. G2 has one old incentivized review. Trustpilot direct numeric is excluded and G2 n=1 is below the minimum external sample; missing panel coverage is not negative evidence. ^{8 5}

Ajelix can ingest business files and, on Enterprise, connect agents to organizational systems and execute workflows. The Enterprise page claims RBAC, audit trails, policy enforcement, sensitive-data controls, human escalation for high-stakes decisions and self-host/private-cloud options. ²

The privacy policy states encryption in transit, access controls, security audits, data minimization, no use of user data to train AI models, GDPR/CCPA rights and Google API Limited Use commitments. These are first-party policy statements, not an independent audit. The policy contains generic wording and a residual placeholder in one contact passage, reducing transparency confidence. Buyers should verify security certifications, subprocessors, DPA/retention/incident terms and exact model-provider routing. ⁹

Because workflows may act without supervision, least privilege, approval gates, rollback/reversibility and logging should be calibrated to blast radius. The vendor describes human escalation for high-stakes decisions; AiToolMap treats that as a design claim, not a safety guarantee. ²

Best suited to teams repeatedly turning business data into deliverables and wanting reusable outputs rather than disposable chat sessions. Enterprise is relevant where governance, RAG and deployment control matter. Most public user evidence still describes the earlier spreadsheet/BI lineage.