Building an enterprise context layer on your data lake
How DevRev connects Snowflake, SAP, ServiceNow, and Oracle into one intelligence surface for instant insight.
WBD
Overview of this document
Following our conversation, here's a comprehensive view of the platform, architecture, and a production case study for your team's review.
Platform architecture
How Computer connects to enterprise systems, builds context, and reasons across data — without importing it.
Production case study
India's largest airline — a 38-skill agent orchestrating payments, rebooking, and operations across 4 enterprise systems in real-time.
Your use cases
How the same pattern applies to WBD — content analytics, vendor intelligence, and finance self-serve via Snowflake, SAP, and ServiceNow.
Proposed next steps
A concrete 4-week POC path — from scope alignment to live validation.
This deck is designed to be self-contained — share internally with stakeholders who need the technical context.
Data is everywhere — insight is nowhere
Your data team navigates 10+ systems to answer one business question. That shouldn't take days.
Fragmented data
Business intelligence spread across Snowflake, SAP, ServiceNow, Oracle ERP, and documents — no unified view.
Manual insight generation
CFO needs content ROI numbers? That means navigating 4 dashboards, 3 applications, and a week of analyst time.
Trapped knowledge
Vendor MSAs, SOWs, contracts, and RFPs hold critical intelligence but aren't searchable or analyzable at scale.
What if one intelligent layer could sit above all your data and answer any question in seconds?
Data in. Intelligence out. Decisions accelerated.
Computer unifies your systems into one permission-aware memory, then reasons and acts across all of it — governed, observable, auditable.
Three problems, one context layer
Each maps directly to DevRev Computer's capabilities — no custom development required.
Email → invoice generation
Email arrives with vendor request → agent looks up opportunity in Salesforce, pulls contract terms from SAP/Ariba, generates invoice — no manual data gathering.
Opportunity enrichment across 5 ERPs
Query a Salesforce opportunity and get meeting transcripts, OneDrive attachments, Ariba procurement history, and Workday team data — unified context without ETL.
Shared context layer for all agents
Multiple vertical agents (SAP-native, Salesforce-native, Ariba) can query one gold-standard context layer — no duplicated integrations, shared knowledge across teams.
India's largest airline — Customer Payment Desk Agent
A fully agentic system orchestrating across 4 enterprise systems via API calls — no data imported into a knowledge graph. Pure context engineering.
38
Skills orchestrated by a single agent
5-10m
Avg handle time (down from 30-45 min manual)
4
Enterprise systems connected via real-time APIs
<5%
Error rate (down from 15-20% manual process)
The agent uses ticket fields as session state, skills make real-time API calls, and the LLM orchestrates the flow — no ETL, no data lake sync needed.
Orchestrate without importing data
Ticket fields store API Tokens, PNRs, order IDs — the agent never asks the human to repeat context.
Zero ETL. Every skill calls the source system's API in real-time with live credentials.
Skills handle deterministic logic (fare math, ancillary codes); the LLM handles flow orchestration and UX.
Full payment desk automation — 38 skills
Rebooking & flight search
Search, sell itinerary, save passengers, replicate ancillary services and seats — all automated across booking engine APIs.
Payment orchestration
Generate payment links, poll status, handle refund verification, fare overrides with tax calculation.
Case lifecycle
Ticket intake, status updates, email notifications, and auto-close — full loop from open to resolved.
Hold release, date mod & refund verify
Release held PNRs, modify travel dates, verify refund status by PNR or transaction, manifest lookup for lost PNRs.
Ask a question. Get a cited answer.
Computer reasons across your entire data estate — not just one database at a time.
Production cost: $1.2B across 8 titles · Revenue: $7.7B theatrical + $2.1B streaming
Blended ROI: 7.2× — sourced from 2 systems
Connects to your existing stack
No rip-and-replace. Computer layers on top with secure, read-only connectors.
Snowflake
Data lake queries, content analytics, engagement metrics
Native connector
SAP
Financials, procurement, project costs, vendor master
Native connector
ServiceNow
IT operations, workflow data, request management
Native connector
Oracle ERP
Financial planning, budgets, actuals, GL data
Native connectorSharePoint, Google Drive, Confluence (for documents) · Custom APIs via universal connector · SSO & RBAC inherited from your IdP
From here to live POC in 4 weeks
A lightweight pilot scoped to one high-value use case — proving value before broader rollout.
This week — align on scope
Pick 1 use case (e.g., content franchise analytics). Define 5-10 representative queries. Identify the data sources needed.
Week 2-3 — connect & configure
DevRev SE team connects Snowflake + one ERP source. Build the knowledge graph. Configure permissions and test queries end-to-end.
Week 4 — validate & expand
Your data team tests real queries. Measure time-to-insight vs. current process. Decision point: expand to additional use cases or full rollout plan.
Let's build together.
One context layer. Every data source. Instant enterprise insight.