.png)
Sparkflows sits on top of HPE Private Cloud AI as your visual agent design and execution layer — so any enterprise team can build intelligent agents, and HPE’s private AI factory runs them at scale, without data ever leaving your organization.
WHY SPARKFLOWS ON HPE
End-to-End Self-Serve AI — on Infrastructure You Control
Sparkflows extends Databricks with visual, self-service tools for building AI Agents, Agentic AI Apps, advanced analytics, and data workflows — making advanced AI and analytics accessible across teams.
.png)
End-to-end self-serve
Design, deploy and scale AI and Gen AI workflows effortlessly on HPE Private Cloud AI.
.png)
Deep integration
Streamlined data analytics and ML workflows across the full PCAI stack.
.png)
Governance & security
Data stays within the organization, with robust governance and security controls.
.png)
Lower total cost of ownership
Low-code / no-code consolidates tools and reduces specialist bottlenecks.
Business User
.png)
Data Analyst
.png)
Data Engineer
.png)
Data Scientist
.png)
AI Engineer
HOW IT WORKS
Three Steps from Idea to Production Agent on HPE Private Cloud AI
Sparkflows manages the full journey — from visual agent design, to NVIDIA NIM–powered reasoning, to governed deployment on your own GPU infrastructure.
DESIGN
Build agents visually, not in code
Drag and drop 500+ nodes — connectors, data prep, ML models, LLM steps and reasoning — onto the Agent Builder canvas. Start from 200+ pre-built agents and templates from week one.
ENRICH
Ground agents in your private data
Pipe structured and unstructured data from the lake into embeddings and vector stores, add ML predictions, and call NVIDIA NIM models — all on-prem, with data quality and lineage tracked end to end.
DEPLOY
Ship to production on HPE PCAI
Publish agents, chat assistants and analytical apps to your HPE Private Cloud AI environment — scheduled, orchestrated and governed, with enterprise access controls and audit built in.
.png)
Agent Builder
LangGraph-based agents
.png)
Agent Orchestrator
Multi-agent workflows
.png)
Chat Assistants
Knowledge-base builder
.png)
Pipeline Designer
Agent Builder
.png)
AutoML & MLOps
Train, track, deploy
.png)
Prompt & LLM Studio
Prompts, RAG, evals
.png)
Data Quality Studio
Profiling · validation
.png)
Reports Designer
Dashboards · what-if
Notebooks
Full-code when needed
UNIFIED WORKBENCH
Anyone can Build.
HPE Runs It.
Business users, analysts, data engineers, data scientists and AI engineers work in one visual interface — no-code, low-code or full-code. Sparkflows routes each workload to the right engine on HPE Private Cloud AI: Spark for data, NVIDIA GPUs for ML and inference, NIM for LLMs.
​
-
No back-and-forth between data teams and IT
-
15–20X more users enabled across functions
-
Consolidate data workflows on one platform
DEEP INTEGRATION
Deeply Integrated with Every Layer of HPE Private Cloud AI
A tightly connected stack from data to model to deployment — and Sparkflows supports all configurations of the HPE Private Cloud AI stack.
MODELS + LLM
NVIDIA NIM
Call NIM microservice models directly from LLM nodes for reasoning, generation and decision steps — inference stays on your GPUs.
.png)
COMPUTE + GPU
NVIDIA AI Enterprise & GPUs
Run ML training, fine-tuning and inference on the NVIDIA-accelerated HPE ProLiant compute inside Private Cloud AI.
PLATFORM + CONSOLE
HPE AI Essentials
Sparkflows runs inside the PCAI software stack, alongside the tools and unified console teams already use to deploy and govern AI.
STORAGE + DATA LAKE
HPE Data Fabric & Alletra
Read and write S3-compatible object storage on-prem or hybrid, with Delta Lake, Apache Iceberg and Apache Hudi support.
.png)
BIG DATA + SCALE
Spark & AI Compute
Push high-volume ETL and feature engineering onto high-performance Spark clusters — scaling to petabytes within agent pipelines.
.png)
STORAGE + RETRIEVAL
Vector DBs & RAG
OCR, chunking, embeddings and vector-DB nodes turn documents into a private knowledge warehouse for RAG and chat.
.png)
MANAGEMENT + OPS
HPE GreenLake Cloud
Operate the private AI factory with a cloud experience — provisioning, monitoring and lifecycle managed through GreenLake.
SECURITY + GOVERNANCE
Enterprise Governance
Data stays inside your organization. Role-based access, encryption, lineage and audit — built for regulated industries.
NEW · HPE DISCOVER 2026
Ready for the Agentic AI Factory
HPE and NVIDIA are taking agentic AI into production on Private Cloud AI. Sparkflows gives every team a no-code way to put that new capability to work.
NVIDIA Agent Toolkit & Nemotron
Open Nemotron models and secure agent runtimes arrive on PCAI — ready to power the reasoning steps in Sparkflows agents.
HPE Alletra Storage MP X10000
AI-ready object storage with metadata automation — HPE cites up to 20X faster token response for RAG and agent workloads.
.png)
HPE Data Fabric Software
Expanded Apache Airflow support and an enterprise AI inventory — a unified data layer for Sparkflows pipelines across edge, core and cloud.
ProLiant Gen12 & Blackwell
ProLiant DL380a Gen12 and DL394 Gen12 with NVIDIA RTX PRO 6000 Blackwell GPUs, with multi-node inference scaling to 256 GPUs.
.png)
Confidential & sovereign AI
NVIDIA Confidential Computing and BlueField zero-trust enforcement for on-prem and sovereign deployments of your agents.
HPE Zerto & OpsRamp
Continuous data protection that flags rogue agent actions, plus hybrid IT operations monitoring across the AI estate.
AGENT & APP STUDIO
Anyone can build.
Azure runs it.
​One canvas for your whole team — Sparkflows routes processing to Fabric, Azure ML, and OpenAI automatically. Design, enrich, deploy, all from one place.
Drag-and-drop canvas
Azure OpenAI
Fabric Lakehouse
No code required
Agentic app builder
AKS deployment
BETTER TOGETHER
A Complete Data Lake Solution
HPE provides the enterprise infrastructure. Sparkflows provides the data engineering, ML and AI layer. Together — an end-to-end solution from raw ingestion to AI-driven insights.
HPE · THE INFRASTRUCTURE LAYER
Data Lake Storage
Scalable object storage — S3-compatible, on-prem or hybrid
HPE PCAI
Unified data platform across cloud, edge and core
Compute Infrastructure
High-performance Spark and AI compute clusters
Security & Governance
High-performance Spark and AI compute clusters
Open Formats
Delta Lake, Apache Iceberg and Apache Hudi support
SPARKFLOWS - THE AGENTIC LAYER
Visual ETL Pipelines
Low-code ETL and transformation on top of the data lake
60+ Connectors
Ingest from any source into the lake without custom code
ML & AutoML
90+ algorithms, experiment tracking and model deployment
Agent Builder
AI agents and orchestration for intelligent automation
Data Quality & Lineage
Profiling, validation and end-to-end data lineage
20X
Faster use cases & analytics
Visual no-code design dramatically shortens the path from idea to a production-ready agent, model or app.
15X
More users enabled
Business users, analysts, engineers, data scientists and AI engineers all build on the same governed HPE environment.
100+
Vertical AI & Gen AI solutions
Ready to deploy and customize — across finance, healthcare, manufacturing, CPG, telecom and more.
.png)
60+
Data connectors
200+
Agents & templates
.png)
90+
ML algorithms
.png)
500+
Gen AI workflow nodes
20+
File formats
.png)
PB+
Scale
USE CASES
What Teams Build with Sparkflows on HPE
From operational automation to Gen AI on private documents — across every enterprise domain.
Fraud Detection
BFSI
Credit Decisioning
BFSI
Patient Risk Stratification
HEALTHCARE
Clinical Trial Optimization
PHARMA
Demand Forecasting
CPG / RETAIL
Customer Churn Prediction
TELECOM
Network Fault Analysis
TELECOM
Quality & Defect Detection
MANUFACTURING
Route & Inventory Optimization
SUPPLY CHAIN
Invoice & Document Intelligence
GEN AI
NLQ Database Reporting
GEN AI
Regulatory Compliance
GEN AI