FRONTFACE YOUR DIGITAL WORKSPACE TWIN
Connect business knowledge, carry memory across tools and AI agents, and turn repeated work into reusable skills with deterministic access controls.
Business knowledge from Salesforce, data warehouses, and documents; conversations from Slack, Teams, and meetings; personal and project memory; OpenAI, AI agents, developer tools, and reusable skills. All connect through deterministic access controls to form FRONTFACE, your digital workspace twin.
Follow one question through all three.
Example question: What was Q3 revenue for Acme?
IllustrativeExample question: What was Q3 revenue for Acme?
IllustrativeProject · Q3 customer review
Applied to querymetric = net_revenue
Q3 is reported as net revenue, not gross.
Hands onmetric = net revenue
Project · Q3 customer review
Applied to querymetric = net_revenue
Three records, one customer.
Hands oncustomer = Acme · 3 sources
Join pathcustomersordersinvoices
MetricNet revenue · governed definition
Each person sees only what their role allows.
Returns3 answers, one per role
WHERE region = 'West' · MASK tax_idAllowedWest accounts only. Tax ID masked.
Every answer passes through all three, on your live data.
See how it worksYour sources
FRONTFACE
Shared capabilities
Who uses it
Five model vendors, 500 questions each, and two conditions. Stating the rule in the prompt produced 34 unauthorised results in 895 scored queries. Enforcing it on the generated query produced none.
Read the analysisAn independent answer key, a negative control, nine seeded faults, and a comparison with PostgreSQL’s own access controls. Four checks that make the measurement open to scrutiny.
Read the analysis147 access roles, 17 rule shapes, and 10 SQL structures. Measuring the combinations behind the query count revealed three defects in our own rewriter.
Read the analysisStart with a workflow your team knows.
Explore these example scenarios.
Bring current results, planning documents, and previous review decisions into one preparation workflow.
“Compare this month’s revenue with our plan and draft the review brief using the assumptions from our last meeting.”
Connect incident records, runbooks, and project decisions to help your team investigate with the relevant history.
“Find related incidents and the approved runbook, then draft a response plan using what we learned last time.”
Bring approved procedures, operational information, and project commitments into a clear team handoff.
“Prepare the operations handoff from our approved procedures and the outstanding commitments from the last team meeting.”
Bring a workflow from your team. Explore how FRONTFACE connects business knowledge, working memory, and your tools within the access rules you define.
Frequently asked questions
How your digital workspace twin fits your knowledge, your work, and your connected tools.
Talk to the teamFRONTFACE is a connected view of your business knowledge, working memory, and reusable skills. It brings mapped business data, document evidence, and relevant memory into the work humans and AI agents do, within their access rules.
The structured-data path establishes a map of schemas, business definitions, and entity relationships, then queries business data at its source. The map is reused across requests and can evolve as sources change. Document retrieval has a separate index of passages, and working memory has its own capture and storage path. These are distinct from copying a source database.
FRONTFACE is designed to serve memory to supported external tools and AI agents through interfaces such as MCP. It can also work through connected applications from its own workspace. Available capabilities depend on the integration: permission to read, to capture activity, and to perform an action are separate scopes.
The platform is designed to connect databases, business applications, and files. Examples include Snowflake, PostgreSQL, Salesforce, spreadsheets, and PDFs. The sources available in your deployment depend on the connectors configured for it. We can review your systems and integration requirements in a demo.
Document retrieval finds relevant evidence in indexed files. Working memory retains what was learned from permitted activity, such as a preference, a project decision, or a lesson from a previous task. Memory records its source and time; a later revision can supersede a fact without erasing its history.
Personal memory belongs to the individual. Project memory is shared with humans who have access to that project. Private material is not automatically promoted into shared memory. In an enterprise deployment, permissions are checked when memory is recalled, just as they are when data is queried.
Deterministic access controls use defined roles, permissions, and data scopes to issue an access pass, verified wherever protected information is used. The AI model cannot expand those permissions. In an enterprise deployment, revoked access is excluded from subsequent queries and memory recalls.
FRONTFACE is designed to return answers with supporting sources, citations, and the scope of the analysis. It can combine structured records with relevant document passages. When a question or its evidence is ambiguous, the response should ask for clarification or make the limitation clear.
A recurring, successful procedure can become a proposed skill. You review and approve it before it is saved for reuse. Personal skills and shared project skills remain appropriately scoped; a procedure learned from private memory is not automatically shared with a team.
That depends on your deployment, connectors, and model provider. Personal memory is designed for local storage, while the enterprise knowledge foundation is designed for infrastructure your organization controls. Cloud connectors and hosted models may process data outside that infrastructure. We can review the complete data flow with your team before choosing a setup.