Back to the Rourkela Steel Plant pitch    Open the pitch book    Download POC checklist (Excel)
SCIKIQ · Account Brief

Rourkela Steel Plant — account brief & discovery

The working notes behind the pitch: where they are on the maturity curve, who's in the buying group, the questions to ask, and how we're positioned against the alternatives.

Internal · for the account team
The thesis

Why Rourkela Steel Plant, why now

Account thesis

Rourkela Steel Plant (RSP), as SAIL's flagship integrated steel unit, is at a pivotal moment: the government has mandated a near-doubling of capacity to 9.8 MTPA by 2030, with a sharp focus on specialty steels, operational excellence, and digital transformation. Recent partnerships (e.g., with ABB for digital twins) and management changes (Alok Verma as Director In-charge) underscore a drive to modernize legacy operations, boost productivity, and reduce import dependence. However, RSP still faces entrenched data silos across production, supply chain, finance, and compliance—hampering real-time visibility, predictive maintenance, and agile decision-making. SCIKIQ can directly unlock value by connecting and contextualizing these silos, accelerating digital initiatives, and giving RSP a data/AI edge over domestic and international rivals as it scales.

Why SCIKIQ for Rourkela Steel Plant — the proof that lands
  • 85% faster data integration: Rapidly unify data from legacy OT/IT systems (blast furnaces, SAP, MES, LIMS, etc.) for a plant-wide 360 view.
  • 70% lower data-prep cost: Automate contextualization of production, quality, and supply chain data for analytics and reporting.
  • 5x faster time-to-market for data products: Enable new digital twins, predictive maintenance, and yield optimization use cases.
  • 95% fewer compliance violations: Strengthen audit, safety, and environmental reporting with end-to-end lineage and governance.
Maturity

RSP is at Stage 2 (Enterprise 360): siloed reporting, pilot digital twins—ready to leap to graph reasoning and agentic automation.

From silos and dashboards to autonomous execution. Our read of Rourkela Steel Plant's current stage is highlighted.

Stage 1

Reporting & Silos

Data is fragmented across OT/IT systems (SAP, MES, LIMS, SCADA); analytics are manual and backward-looking.

  • Departmental Excel/Access reports dominate decision-making.
  • No unified plant/production 360; root-cause analysis is slow.
  • Manual compliance and audit trails.
Likely today
Stage 2

Enterprise 360

Initial data integration projects (digital twins, ABB partnership); some unified dashboards, but limited cross-functional reasoning.

  • Pilot digital twins for blast/oxygen furnaces.
  • Some central dashboards, but not real-time or contextual.
  • Ad hoc data science/AI pilots, not at scale.
Stage 3

Reasoning: Graph + Copilot

Contextualized knowledge graph links assets, process, quality, and commercial data; semantic search and plain-language AI copilots.

  • Engineers can query root causes and scenarios in plain language.
  • Incident paths and dependencies are mapped for rapid troubleshooting.
  • Cross-functional AI/ML models leverage unified data.
Stage 4

Autonomous: Agents

Autonomous agents detect, recommend, and execute corrective actions (e.g., order re-routing, predictive maintenance, compliance filings) with write-back to core systems.

  • Closed-loop automation of key workflows (maintenance, quality, supply chain).
  • Agents post transactions directly to SAP/MES.
  • Quantified impact on yield, cost, and compliance.
Stakeholder map

Who's in the room — and the line that lands

The buying group for an enterprise-AI platform, with each persona's concern and the message that resonates.

Director In-chargeeconomic buyer
Shri Alok Verma
Cares about: Meeting 2030 capacity and productivity targets, reducing cost/ton, and digital transformation success.
“SCIKIQ will accelerate your modernization mandate, unify plant data, and deliver a step-change in productivity and compliance.”
Executive Director (Projects/IT)champion
Cares about: Successful delivery of digital twin, automation, and analytics projects; integration of legacy and new systems.
“SCIKIQ de-risks and speeds up integration—no-code, 200+ connectors, and proven in heavy industry.”
General Manager (Production/Operations)user
Cares about: Real-time visibility into production, downtime, and yield; actionable insights for OEE and throughput.
“With SCIKIQ, you get a live, contextualized plant 360—see, explain, and fix issues before they hit output.”
General Manager (Quality/Process)user
Cares about: Reducing defects, root-cause analysis, and compliance with customer specs and environmental norms.
“SCIKIQ's knowledge graph and copilot let you trace, explain, and automate quality interventions—faster and with full audit.”
Chief Financial Officereconomic buyer
Cares about: Margin, cost-to-income ratio, working capital, and regulatory compliance.
“SCIKIQ lowers integration and compliance costs, improves cash conversion, and reduces risk of audit findings.”
Head of Digital Transformation / ITchampion
Cares about: Delivering on digital strategy, de-risking legacy, and showing quick wins.
“SCIKIQ is a true AI-first data fabric—no-code, fast to value, and proven in complex manufacturing.”
CISO / Head of Complianceblocker
Cares about: Data security, access controls, and regulatory reporting.
“SCIKIQ brings end-to-end lineage, fine-grained access, and automated compliance reporting out-of-the-box.”
Discovery

Questions to ask in the meeting

Data & context

  • Which core OT/IT systems (SAP, MES, LIMS, SCADA, etc.) are most critical and most siloed?
  • Where do you lack real-time or contextualized data for decision-making?
  • How is data currently shared between production, quality, and commercial teams?

Digital transformation & strategy

  • What are the top 3 digital initiatives tied to the 2030 capacity and productivity targets?
  • How do you measure success for digital twin and automation projects?
  • Where have past integration efforts stalled or overrun?

Operational pain points

  • Where do production delays, downtime, or quality issues most impact output and margin?
  • What is the current process for root-cause analysis and incident response?
  • Which workflows are most manual and ripe for automation?

Compliance & governance

  • How do you ensure data lineage and auditability for safety, environmental, and financial reporting?
  • Where are compliance violations or audit findings most frequent?
  • What are the biggest risks in regulatory reporting today?

AI & analytics

  • What AI/ML use cases have shown the most promise (predictive maintenance, yield optimization, etc.)?
  • How do you operationalize AI models—are they embedded in workflows or still in pilot?
  • Would a plain-language copilot or autonomous agent change how your teams work?
Competitive landscape

RSP faces a crowded field: legacy IT, new digital twins, and global data/AI platforms—SCIKIQ wins on speed, context, and agentic automation.

Rourkela Steel Plant is evaluating a mix of legacy vendors, digital twin partners (ABB), and horizontal data/AI platforms. The main alternatives are large-scale platforms (Palantir, Databricks), Microsoft's new Fabric, generic data fabrics, and industry-specific analytics/automation tools. SCIKIQ's edge is its no-code, AI-first contextualization engine, rapid integration of legacy OT/IT, and ability to move beyond dashboards to agentic execution—critical as RSP scales and digitizes.

Palantir Foundry
Global industrial data platform, strong in process/asset graphs and digital twins.
SCIKIQ edge: SCIKIQ is faster to deploy (no-code), more cost-effective, and natively agentic—no need for custom builds.
Databricks
Lakehouse platform, strong for data science and ML on unified data.
SCIKIQ edge: SCIKIQ brings contextualization, knowledge graph, and agentic automation out-of-the-box—beyond raw data lakes.
Microsoft Fabric
Unified analytics fabric, deep integration with Microsoft stack.
SCIKIQ edge: SCIKIQ is system-agnostic, with 200+ connectors and manufacturing-specific lineage, governance, and agent frameworks.
ABB Digital Twin / Automation Suite
Industry-specific digital twin and automation tools, focused on OT integration.
SCIKIQ edge: SCIKIQ complements ABB by connecting OT/IT/ERP, contextualizing data, and enabling cross-functional AI and agents.
Build-it-yourself (internal IT/consulting)
Custom integration, dashboards, and AI pilots built in-house or with SIs.
SCIKIQ edge: SCIKIQ delivers 85% faster integration, 70% lower prep cost, and proven compliance—at a fraction of the risk and time.
Niche graph/semantic vendors
Specialized graph or semantic layer tools, often lacking scale or agentic automation.
SCIKIQ edge: SCIKIQ is enterprise-grade, proven in manufacturing, and delivers end-to-end: from graph to copilot to agentic execution.
POC requirements

How we'd prove it — the ScikIQ POC, layer by layer

Download checklist (Excel)

A POC proves ScikIQ's feasibility against Rourkela Steel Plant's data needs — installed, configured and tested inside your environment to validate a set of business, functional, technical and operational goals. Every POC covers three things: technical & functional validation, deployment sizing, and ROI.

Problem statement & financial driver — revenue or cost; regulatory or discretionary spend.
Key success criteria (KPIs) and decision criteria — technical, economic and benchmarking.
Risks — organizational/political, technical, commercial — and the named economic buyer.
01

Enterprise 360

ScikIQ Data Integration · Connect

Connect Rourkela Steel Plant's structured & unstructured sources and build the unified Business 360 with no-code pipelines — cutting data-to-action from months to days.

Validate in POC
Scope inputs needed
Success criteria
Applicable SKUs
SCIDI001 · Document (Mongo DB)SCIDI002 · Real-time / StreamingSCIDI003 · BatchSCIDI004 · SAPSCIDI005 · Log-based CDCSCIDI006 · API
ScikIQ POC Guide — Data Integration POC
02

Knowledge Graph

ScikIQ Data Governance · Knowledge Graph & Lineage

Model Rourkela Steel Plant's entities and relationships into a living knowledge graph with end-to-end lineage, cataloguing and quality — so AI can traverse cause → effect.

Validate in POC
Scope inputs needed
Success criteria
Applicable SKUs
SCIDGI001 · Data CatalogSCIDG002 · Metadata DiscoverySCIDGI003 · Asset Approval & Search (Elasticsearch)SCIDGI004 · Knowledge Graphs (Neo4j) & Data LineageSCIDGI005 · Data Quality & Data Observatory
ScikIQ POC Guide — Data Governance POC
03

AI Copilot

ScikIQ GenAI Studio · Talk to your data

Ground a conversational copilot on Rourkela Steel Plant's knowledge graph + semantic layer — plain-language operational, commercial and risk queries with explainable, auditable answers.

Validate in POC
Scope inputs needed
Success criteria
Applicable SKUs
SCIAI001 · GenAI Studio — Conversational CopilotSCIAI002 · Semantic Search (structured + unstructured)SCIAI003 · Grounding & Explainability (graph-RAG)SCIAI004 · Guardrails & Governance for GenAI
Authored to the POC Guide structure (step not in the source doc)
04

Agent Factory

ScikIQ Agent Factory · No-code autonomous agents

Build no-code agents that act on Rourkela Steel Plant's live context — detect, reason and close the loop with a real transaction in the source system, under human-in-the-loop guardrails.

Validate in POC
Scope inputs needed
Success criteria
Applicable SKUs
SCIAG001 · No-code Agent BuilderSCIAG002 · Triggers & OrchestrationSCIAG003 · Closed-loop Connectors (IT/OT write-back)SCIAG004 · Agent Governance, Approvals & Audit
Authored to the POC Guide structure (step not in the source doc)
POC readiness checklist
Kick-off
Data readiness
IT readiness
Testing readiness
Battle card

Objection handling — across all four layers

Field-ready objection handling for Rourkela Steel Plant, layer by layer — grounded in the SCIKIQ Battle Cards. For each: the objection you'll hear, the response that wins it, the proof, and who you're really competing with.

Buyer: C-suite (CIO, CTO, CFO) and leaders in data, compliance and innovation.
01

Enterprise 360

Data Hub & Lakehouse · Innovation at speed
“We're happy with our current data stack and tools.”
We complement and enhance what you have — no rip-and-replace. One no-code platform unifies all data across cloud/hybrid and adds AutoML & GenAI value your current stack can't reach.
“We already have a data lake / warehouse.”
Separate lakes and warehouses raise cost and slow real-time analytics. SCIKIQ unifies them and builds the Business 360 on top — no data movement.
“Our SI / in-house team can build it.”
That's years of pipelines and heavy services spend. SCIKIQ delivers strategy-to-execution on one platform — up to 80% cost savings, <6 months to value, 200+ no-code connectors.
“Another integration project that stalls in IT.”
No-code pipelines move integration to the business team; data-to-action drops from months to days — proven on your data in the POC.
200+ connectors · no data movementUp to 80% cost savingsForrester Top-34 augmented-BINo-code · <6 months to value
Real competition: Big-4 & boutique data firms (strong on strategy, light on execution), global / local SIs (vendor-tied, generalized, services-heavy), plus Informatica/Fivetran & build-it-yourself. Wedge: one no-code platform, strategy-to-execution — a Business 360, not just pipes.
02

Knowledge Graph

Data Governance · Governance on autopilot
“We already have a data-governance solution.”
We enhance rather than replace — a no-code, metadata-first, GenAI-integrated layer that boosts your governance and builds the knowledge graph + lineage on top.
“A BI dashboard already shows what's happening.”
Dashboards answer what; only a graph answers why. Typed relationships + column-level lineage let AI traverse cause → effect across silos.
“Can it scale to our complex cloud / hybrid data?”
A modular, flexible architecture adapts to growing volumes and new sources across complex cloud/hybrid stacks, continuously updated with the latest tech.
“How do we trust the relationships?”
Every edge is lineage-traced and governed; GenAI authors the rules (manual rule creation is ~70% slower) — fewer errors, lower cost to maintain.
Graph + lineage pre-built (Neo4j)Metadata-first · GenAI rule authoringForrester DG challenger~70% faster rule creation
Real competition: Big-4 & boutique data firms (strong on strategy, light on execution), global / local SIs (vendor-tied, generalized, services-heavy), plus Palantir Foundry & niche graph vendors. Wedge: governed, metadata-first graph + lineage — no-code and faster to value.
03

AI Copilot

Gen AI · Talk to your data
“Do we really need a GenAI platform? We're good today.”
Chat-based access puts data in everyone's hands and lifts data literacy org-wide. Grounded on your graph, answers are explainable — not generic chatbot guesses.
“We'll just use ChatGPT / a generic copilot.”
Ungrounded models hallucinate on enterprise data. Ours is grounded on your graph + semantic layer with citations and lineage; RBAC honoured in every answer.
“GenAI is still maturing — invest now?”
Every tech matures; our engineers keep the platform current so it never goes stale. Start with one department, prove ROI, then roll out.
“LLMs can't be trusted with our numbers / security.”
Every figure cites its source and path; quality & freshness gate what it answers, and row-level security is honoured inside every answer.
Graph-grounded (no hallucination)Explainable & lineage-tracedChat access · data literacyRBAC enforced
Real competition: raw LLMs/chatbots, BI NLQ, and Big-4 & boutique data firms (strong on strategy, light on execution)' GenAI services. Wedge: graph-grounded, governed, auditable — and democratised access.
04

Agent Factory

Machine Learning & Auto ML · Automate data processes
“We already have AutoML / automation.”
Replace point automation with a holistic no-code platform — more capabilities and value, and agents that close the loop, not just score models.
“Autonomous agents are too risky in production.”
Human-in-the-loop approvals, full audit and safe-stop are built in; agents run in a sandbox first and you own the approval matrix.
“RPA already automates our workflows.”
RPA scripts brittle UI steps; agents reason on live graph context and close the loop via APIs — incident response, compliance, optimization.
“Why now / no special skills on the team?”
Begin today — automation cuts this year's spend itself: no code, no special skills, immediate results. ROI aligns future budgets.
No-code agent builderClosed-loop write-back to IT/OTApprovals · audit · safe-stopNo special skills needed
Real competition: RPA (UiPath), AutoML point tools & bespoke scripts, plus global / local SIs (vendor-tied, generalized, services-heavy). Wedge: context-aware, governed, closed-loop on one platform.
Objections you'll hear at every layer
“No budget / we don't need it right now.”
Begin with a phased pilot on one domain — ROI shows in days and aligns next year's budget. The best firms modernise every year; the competition won't wait.
“Long-term support & reliability?”
Although the platform is no-code, a dedicated support team is always available, with long-standing customer references.