Knowledge
What things mean and how they connect. An active SKU, a recognized dollar, a critical alarm—one governed definition, not three departments arguing over their own.
Definitions · relationships · entitiesContext OS for Telecom + Retail
Structure enterprise knowledge. Give every agent the right truth.
We are building the governed Context OS that connects live business, customer, product, operational, and financial knowledge—making Telecom and Retail agents accurate, contextual, and useful for every permitted persona.
What context actually means
Three things every agent needs before it's useful.
Most "context" conversations stop at data. An agent inside a real business needs three distinct things, and confusing them is why so many agents ship fast and then stall.
Expertise
How work actually gets done here. The steps a senior planner takes without thinking twice, still living in a Slack thread and one person's head.
Procedures · playbooks · know-howNorms
What's allowed, and who has to say yes. Which discount needs approval, which action needs a human in the loop, which data can't leave the region.
Policy · permissions · approvalsMost context layers stop at the first one. This one is built to carry all three.
"Why did the quality hold rate spike on line 3 this week?"
Context typeQuestion it raisesAnswer it needs
Knowledge
What does "quality hold" mean, and where is it recorded?
The certified QMS + MES definition of a hold event—not each supervisor's private tally.
Expertise
How does an experienced quality engineer investigate this?
Check the lot, supplier, and process-parameter correlation before assuming a process drift.
Norms
Who's asking, and what can they see?
A shift supervisor sees their line; a plant quality manager sees the full site; any corrective action still needs sign-off.
Choose the operating domain
Context starts with how your business works.
Select the product closest to the decisions your agents must support. Each page maps the relevant entities, events, systems, roles, controls, and outcome stories.
01 · Industrial operations
Manufacturing
Connect assets, production orders, materials, process conditions, quality genealogy, maintenance, workforce, and supply constraints.
Context layer: Aligns machine events with the order, material genealogy, quality limits, maintenance history, skills, and supply conditions behind the decision.
Explore Manufacturing →02 · Communications
Telecom OSS + BSS
Connect network and service state with catalog, quote, order, activation, usage, billing, revenue, subledger, and GL context.
Context layer: Joins resource and service topology to customer promises, quote-to-order, usage, billing, assurance, revenue recognition, and GL evidence.
Explore Telecom →03 · Unified commerce
Retail
Connect products, prices, promotions, inventory, shoppers, stores, orders, fulfillment, returns, suppliers, and margin.
Context layer: Carries product and promotion rules through live store and DC inventory, shopper consent, orders, fulfillment, returns, supplier funding, and margin.
Explore Retail →Not everything is semantic search
Route each question to the right kind of truth.
Embeddings help with meaning and similarity. They are the wrong tool for an exact alarm code, a governed KPI, a changing service topology, or a live order state. The Context OS selects and combines the retrieval paths a task actually needs.
Relational + metric
Use deterministic queries for orders, inventory, billing, SLAs, governed measures, and other structured facts where precision matters.
Exact valuesjoinsaggregationsKeyword + lexical
Find exact part numbers, alarm codes, clauses, tickets, product IDs, and specialist terminology without fuzzy substitution.
IdentifiersBM25filtersSemantic + multimodal
Retrieve conceptually related manuals, notes, procedures, images, and unstructured knowledge when the wording does not match.
EmbeddingsrerankingmetadataGraph + topology
Traverse how assets, services, resources, customers, products, and dependencies connect across operational domains.
Entity resolutionpathsimpactEvent + temporal
Reason over current state, event order, validity windows, changes, and prior decisions instead of treating every fact as timeless.
Streamstime seriesmemoryTool + action context
Call authoritative APIs when an answer must be checked live, then expose approved actions with identity, policy, and human gates.
MCPOpenAPIworkflow toolsContext router
Classify the task → resolve identity and entitlements → choose one or more retrieval paths → rerank and reconcile evidence → return the smallest useful context package with provenance.
Engineered, not assembled
Depth becomes product decisions.
Context engineering, retrieval, memory, and agent security are architecture decisions made here, not integration afterthoughts—validated against real accuracy and governance requirements before anything ships.
Context is a system, not a longer prompt.
Retrieval, processing, and memory are managed as one governed layer that spans every agent and tool—not reassembled by hand inside each prompt.
Enterprise truth needs hybrid retrieval.
Vector similarity and graph traversal work together, so a question about relationships and a question about meaning both get answered from the same governed layer.
Memory should reconcile, not silently accumulate.
Episodic events stay separate from stable knowledge, with conflict-aware reconciliation—so a correction updates the record instead of just piling on top of it.
Agent identity must survive every handoff.
Identity, delegation, revocation, and traceability are treated as infrastructure that travels with the request—not a one-time login check at the front door.
How we validate: Every architecture decision is checked against domain-specific accuracy, permission, freshness, cost, latency, and auditability requirements before it reaches the product.
The architecture
A context fabric, not another data copy.
The layer federates existing systems, chooses the right retrieval method, and packages the smallest useful evidence set for the task. Materialized indexes are selective; authoritative state stays authoritative.
Federated truth
SQL · lakehouse · metrics
Documents · tickets · media
Knowledge + topology graphs
Streams · telemetry · live APIs
BodhiContextAI
Context router + policy plane
Intent, entity + persona resolutionHybrid retrieval + graph traversalTemporal assembly + scoped memoryPolicy, provenance + evaluationTool contracts + approval gates
Context contracts
REST · GraphQL · SQL
MCP · OpenAPI tools
Events · agent-to-agent handoffs
Answers · actions · evidence
Where this fits
Your systems stay authoritative. We make them legible to agents.
MES, ERP, OSS/BSS, CRM, and whatever data platform you already run govern their own slice well. What none of them does natively is turn that technical governance into business meaning any agent can act on across all of them at once.
Technical governance, one system at a time.
Access control, schema, and lineage inside each system are often solid on their own—that isn't the gap. The gap opens at the boundary, where an agent needs one fact that spans systems built to guard their slice, not share it.
Business meaning, served to any agent, via MCP.
One governed definition per concept, cross-system lineage, and policy that travels with the data—delivered through open MCP so any agent, not just one built for a single platform, can act on it.
Skip this layer, and each system just gets a smarter interface bolted onto the same old boundary—context locked to one platform, one team, one slice of the business.
What this is not
Adjacent, but not the same thing.
Not a data catalog
A catalog helps people browse tables. This is built for agents to act on—context is exposed through MCP and APIs, not a search box for analysts.
Not a semantic layer
A semantic layer defines what a metric means. This adds who's allowed to see it, how it changed, and what to do about it—across every system, not just your warehouse.
Not chatbot memory
Memory here is scoped, versioned, and governed by precedence rules—not a growing pile of chat history one agent quietly hoards and no one else can check.
Persona-aware context
The same event should not produce the same view for everyone.
An operator, reliability engineer, plant manager, network engineer, care agent, and finance leader have different responsibilities and permissions. BodhiContextAI is being designed to assemble context and expose actions for the authenticated persona and the purpose of the task.
Who is asking?
Bind workforce or customer identity to tenant, role, team, site, region, entitlements, delegation, and session risk.
SSO · IAM · RBAC · ABACWhat may they know?
Apply row, field, document, entity, geography, and purpose restrictions before retrieval results reach the model.
Masking · filtering · purposeWhat helps this persona?
Select the vocabulary, depth, evidence, workflow history, and recommendations appropriate to the person’s job and current task.
Role view · preferences · memoryWhat may they do?
Separate read from action privileges, enforce tool allowlists and limits, and require human approval for high-impact changes.
Guardrails · approvals · least privilegeExample personaContext this role may receiveContext and actions denied
Product management
Approved, aggregated, and de-identified product signals; roadmap feedback; adoption trends.
Raw plant telemetry, network alarms, customer PII, detailed topology, and every operational control tool.
Plant operator
Assigned line state, active order, relevant SOPs, quality limits, and local shift history.
Other sites, unrestricted cost or HR data, and engineering or maintenance actions outside the operator’s authority.
NOC engineer
Authorized domains, alarms, topology, service impact, change windows, and runbooks.
Unmasked billing or customer data, unrelated regions, and network changes outside the approved control scope.
Customer care
Customer entitlement, service health summary, known incident, order state, and approved resolution steps.
Raw network configuration, sensitive topology, engineering credentials, and direct infrastructure controls.
Real-time context + delta intelligence
Fresh by contract. Reconstructable by design.
“Real time” is not one refresh rate. Machine signals may arrive in milliseconds, orders through change data capture, policies through versioned publication, and manuals on a scheduled sync. Each context product should declare its freshness objective and preserve the change history behind it.
Capture the change
Consume event streams, CDC logs, webhooks, API reads, and scheduled snapshots. Retain source offset, event time, ingestion time, schema version, and actor or service identity.
Kafka · CDC · OPC UA events · webhooksUpdate only what changed
Resolve impacted entities, calculate before-and-after deltas, then incrementally update relational views, search indexes, embeddings, graph edges, caches, and temporal memory.
Incremental index · graph delta · cache invalidationRe-evaluate access
Run policy checks when a record, relationship, entitlement, or classification changes. Revoke stale context and notify affected agents instead of waiting for the next query.
Policy version · entitlement delta · invalidationReplay decision time
Answer what the source said, which version the agent saw, what policy allowed, and what action followed at a specific point in time.
Valid time · system time · immutable audit ledgerChange pathSource event→Entity + policy delta→Affected projections→Context-changed event→Agent refresh or revocation
An evidence trail, not only a chat transcript.
Record the identity and purpose, policy decision, retrieval routes, source versions, context supplied, model and tool calls, approvals, action result, and downstream outcome. Sensitive payloads can be redacted while retaining a verifiable control trail.
- Trace IDs across agent, retrieval, and tool execution
- Policy decision and approval records
- Source freshness, lineage, and output citations
- Replayable evaluations and incident review
Learn safely without rewriting truth.
Capture explicit corrections, accepted or rejected recommendations, task outcomes, evaluator scores, and recurring retrieval failures. Promote useful learning into scoped, versioned memory only after validation.
- Personal memory for preferences and working style
- Role and team memory for reusable operating knowledge
- Curated enterprise memory for approved definitions and policy
- Offline evaluation gates, canary rollout, and rollback
Memory precedenceEnterprise policy and authoritative data→Approved team knowledge→Personal preferences→Session context
Lower scopes may personalize presentation and workflow; they cannot override permissions, governed definitions, safety controls, or authoritative operational state.
What we are building
Cloud native by design. Hosted on your terms.
We are building the BodhiContextAI Context OS as modular, API-first cloud infrastructure. It is designed to connect to the systems you already run, use the model stack you choose, and fit the security boundary your organization requires.
Now buildingA governed context foundation for production agents
Pluggable retrievalUse the SQL, search, vector, graph, stream, and cache engines already approved.
Private connectivityConnect systems without exposing private data sources publicly.
Model + protocol choiceUse approved models and expose governed context through APIs or MCP.
Evidence envelopesCarry source, retrieval path, timestamp, policy, confidence, and lineage with every response.
Agent workloads run inside your own virtual network and private endpoints—the same enterprise networking pattern your other production systems already use.
01 · Manufacturing blueprint
Move from plant signals to coordinated action.
Production, quality, maintenance, engineering, and supply teams each see part of an exception. The Context OS blueprint combines OPC UA live state, Asset Administration Shell semantics, MES/ERP transactions, maintenance history, exact identifiers, manuals, and dependency graphs into one governed decision package.
Outcome story · Production exception
A line is losing throughput. What should happen next?
The agent connects the live machine signal with the active production order, recent engineering change, maintenance history, quality limits, material availability, and downstream schedule.
- DetectCycle time is drifting beyond the product-specific control range.
- ExplainA changed component and an overdue calibration are correlated with the deviation.
- EvaluateContinuing risks a quality hold; stopping now affects two downstream orders.
- CoordinateRecommend a controlled maintenance window, resequence orders, reserve the part, and open the quality check—with evidence and approvals attached.
Production exception resolution
Bring schedule, material, asset, labor, and process constraints into one decision instead of reconciling them in a war room.
MESERPAPSWMSMaintenance in operating context
Prioritize work by failure risk, current product run, spare availability, technician skills, and downstream production impact.
IoTEAM/CMMSmanualsinventoryQuality and traceability
Trace a non-conformance through lots, suppliers, process parameters, inspection results, and engineering changes.
QMSPLMgenealogysupplier dataConnected planning
Surface how a demand, capacity, supplier, or maintenance change cascades through production, inventory, and customer commitments.
S&OPplanningprocurementordersStandards-aware direction: Secure industrial information exchange from field to cloud, and a standardized digital representation for assets, are built into how the manufacturing context layer reads the plant floor.
02 · Telecom OSS + BSS
Connect network, commercial, and financial state end to end.
OSS knows what was designed, activated, and changed in the network. BSS knows what was offered, ordered, consumed, charged, billed, collected, and promised. Finance knows what reached the subledger and GL—and what may be recognized. The Context OS links these views without flattening their controls or accounting meaning.
Outcome story · Quote to recognized revenue
A service is live. Can every dollar be explained?
The agent traces the approved quote and contract through product-order decomposition, service activation, usage mediation, rating, invoice lines, credits, subledger entries, GL mappings, and the revenue schedule.
- SellValidate catalog eligibility, negotiated pricing, discounts, approvals, contract terms, and candidate performance obligations.
- FulfillTrace the product order into service and resource orders, dependencies, inventory assignments, milestones, and activation evidence.
- MonetizeReconcile usage through mediation, rating, charging, invoicing, adjustments, collections, and partner settlement; surface leakage and fallout.
- AccountConnect billed and unbilled positions to subledger and GL entries, contract assets or liabilities, and recognition schedules—with policy, lineage, and approval attached.
Shared context chainQuote→Order→Activate→Usage→Charge→Bill→Collect→Recognize→GL
Service + customer assurance
Correlate alarms across domains, map them to services, orders, SLAs, and customers, and prioritize remediation by operational and commercial impact.
NMStopologyservice inventoryCRMCatalog, CPQ + commercial policy
Ground recommendations in product compatibility, eligibility, pricing, discount authority, contract terms, channel rules, and available network coverage.
CatalogCPQpricingeligibilityQuote-to-order orchestration
Preserve the commercial intent as a quote becomes a product order and decomposes into service, resource, partner, and billing orders.
Quoteproduct orderservice orderresource orderFulfilment, activation + inventory
Explain fallout across orchestration, logical and physical inventory, provisioning, number or device assignment, activation tests, and partner dependencies.
Order falloutprovisioningactivationinventoryUsage, charging + billing
Trace usage records through mediation, rating, online or offline charging, balance impact, bill calculation, tax inputs, invoices, credits, and collections.
MediationratingchargingbillingcollectionsRevenue assurance + settlement
Compare what was ordered, activated, used, charged, billed, and paid to detect leakage, duplicate or missing events, margin erosion, and partner-settlement exceptions.
LeakagereconciliationdisputessettlementSubledger + GL reporting
Carry source transaction, account mapping, legal entity, cost center, tax, currency, journal, period, and reconciliation evidence from BSS into finance reporting.
Subledgerjournalschart of accountscloseRevenue recognition context
Connect contract terms, performance obligations, standalone selling prices, allocation, activation milestones, usage, modifications, credits, and deferrals to governed recognition schedules.
ASC 606 / IFRS 15 contextschedulesevidenceAI-native telecom direction: Domain agents, API-exposed OSS/BSS capabilities, intent-driven operations, and a shared temporal knowledge graph combine for cross-domain decisions—not a single system trying to model everything at once.
Start narrow. Compound value.
Prove one workflow, then turn its context into shared infrastructure.
We begin with the decision or action that matters, map the minimum context it needs, establish quality and governance, and expose it through a reusable contract. The next agent starts ahead.
Choose the decision
Define the user, workflow, systems, risk, and evidence that signal a useful outcome.
Build the context slice
Connect sources, resolve meaning, apply policy, and evaluate answers against real cases.
Productize the contract
Serve stable context interfaces, monitor quality, and extend across agents and business units.