Nepal Data Labs
Nepal Data Labs
Ask business data. See the evidence.
Nepal Data Labs builds Dataforge: governed conversational analytics for approved enterprise databases.
AI ProductsData PlatformsGoverned Analytics
Overdue distributors14Applied range: current open invoices
| Distributor | Region | Balance |
|---|---|---|
| Aster Trade | Pokhara | NPR 3.2M |
| Himal Supply | Biratnagar | NPR 2.8M |
| Metro Dealer | Kathmandu | NPR 2.4M |
View query and source
SELECT distributor, region, outstanding_amt FROM approved_credit_view WHERE credit_days > 60;OrganizationNepal Data Labs
Flagship productDataforge
The Organization
AI data products for controlled enterprise work.
Built around structured data, explainable output, private deployment, and real business workflows.
Dataforge
POC first
Flagship Product
Dataforge removes the report queue.
Ask from approved data. Inspect the answer path.
QuestionWhich distributors exceeded their 60-day credit limit?
Ask
Business question.
Map
Terms to schema.
Control
Read-only query path.
Answer
Result with evidence.
Buying Team
Clear for every reviewer.
Answers
Operational questions and follow-ups.
Control
Schemas, metrics, SQL, evidence.
Boundaries
Deployment, access, read-only use.
Decision
POC criteria and rollout fit.
Product Preview
Question in. Evidence out.
Dataforge answer
Three distributors need review based on overdue balance, credit period, and recent payment behavior.
3Priority accounts
Generated read-only query
SELECT distributor, region, risk_level
FROM dealer_credit_view
WHERE overdue_days > credit_days
ORDER BY overdue_balance DESC
LIMIT 10;| Entity | Segment | Priority |
|---|---|---|
| Aster Trade | Pokhara | High |
| Himal Supply | Biratnagar | Medium |
| Metro Dealer | Kathmandu | Medium |
Sourcedealer_credit_view
DefinitionOverdue = outstanding invoice beyond approved credit period.
ControlRead-only access against approved reporting data.
Predicted follow-ups
Problem
The data exists. The answer is stuck.
SQL Barrier
Business users depend on technical teams to retrieve information.
Dashboard Dependency
Dashboards answer predefined questions but cannot anticipate every operational question.
Analyst Bottleneck
Analysts repeatedly prepare routine reports instead of focusing on higher-value work.
Evidence Standard
Every answer needs a trail.
Fast answers are useful. Visible evidence makes them usable.
Original intent
Terms to schema
Generated query
Tables and filters
Authorized answers
Context preserved
Conversation
Follow-ups keep context.
Ask DataforgeShow sales by region
Answer
Top regions ranked by sales for the selected period.
5Regions returned
Sourceregional_sales_summary
SELECT region, SUM(net_sales) FROM regional_sales_summary GROUP BY region ORDER BY 2 DESC LIMIT 5;Business Context
Business words become data logic.
cust_idCustomer
dist_cdDistributor
outstanding_amtOutstanding Balance
sku_qtyInventory Quantity
credit_daysCredit Period
Deployment
SaaS. Private VPC. On-premise.
Managed SaaS
Best for: Mid-market organizations, startups, e-commerce, and digital platforms.
Fast setup and managed infrastructure.
01Managed by Nepal Data Labs
02Good for faster evaluation
03Operational responsibility handled by NDL
Private VPC
Best for: FinTech, enterprise software, logistics, and regulated organizations.
Run Dataforge inside the customer's AWS, Azure, or GCP environment.
01Customer-controlled cloud
02Private infrastructure
03Useful for stronger data isolation requirements
Air-Gapped / On-Premise
Best for: Banks, government, healthcare, and high-security organizations.
Run inside local or private infrastructure without external internet dependency where required.
01Customer data center
02Maximum infrastructure control
03Designed for strict data-sovereignty needs
Security
Enterprise analytics without giving up control.
Read-only query safetyRole-based access controlRead-replica supportData sovereigntyPrivate deploymentAnswer explainability
Solutions
Real questions by industry.
Manufacturing
Manufacturing information is often distributed across ERP, sales, inventory, production, and distributor systems.
Which products are below safety stock?
Banking & Financial Services
Regulated financial institutions need explainable analytics, private deployment, role-based access, and careful data handling.
Show loans overdue by more than 90 days.
FinTech
High-growth financial technology teams need analytics flexibility while preserving data isolation and role-controlled answers.
Compare transaction growth by customer segment.
Healthcare
Healthcare organizations often need controlled analytics for facilities, resources, operations, and service-volume datasets.
Compare service volume across facilities.
Government / Regulated Enterprise
Strict data-sovereignty and infrastructure-control requirements make generic cloud-only analytics difficult to evaluate.
Summarize service volume by region.
POC Readiness
Bring five things.
Dataset
Approved reporting data.
Questions
Real stakeholder asks.
Definitions
Metrics, aliases, dates.
Access
Roles and boundaries.
Success
Accuracy and adoption bar.
Integrations
Connect to the data stack you already use.
Only verified or clearly labeled integration statuses are shown.
PostgreSQLDatabase
AvailableConnect approved relational databases for governed read-only analytics.
MySQLDatabase
AvailableUse structured operational tables as part of a controlled evaluation.
Microsoft SQL ServerDatabase
AvailableSupport enterprise SQL Server environments through approved connections.
OracleDatabase
Custom IntegrationEvaluate Oracle environments through a scoped enterprise integration.
SnowflakeData Warehouse
PlannedPlanned warehouse support; verify roadmap timing before committing.
BigQueryData Warehouse
PlannedPlanned warehouse support for cloud analytics teams.
RedshiftData Warehouse
PlannedPlanned support for AWS data warehouse environments.
DatabricksData Warehouse
PlannedPlanned support for lakehouse-style analytics stacks.
POC
A controlled evaluation path.
Sandbox
Provide a non-sensitive sandbox database or approved sample dataset.
Deploy
Connect data, map schema, add business terminology, and configure permissions.
Validate
Test real operational questions, security behavior, and user acceptance.