ELEVATE NXT 2026 — Round 3
Infrastructure Governance
& Audit Readiness Engine
for Power Distribution Utilities
A read-only governance overlay with explainable AI — detecting cross-system risk before failures occur, making every operational response institutionally defensible, and creating the continuous governance record that transforms how India's DISCOMs relate to their regulators.
Deeptech
Public Infrastructure Governance
Power Distribution
TRL 4 → 6
Karnataka-First
SkyEdgeAI · Bengaluru, Karnataka · Founded 2025 · Grant Ask: Rs. 1 Crore
The Team
Built by People Who Have Done This Before
Domain depth, AI systems expertise, and hands-on power sector experience — in one team.
Afzal Jan
Co-Founder & CTO
AI/ML Systems · Governance Architecture
Architect of multi-domain AI platforms for regulated cyber-physical infrastructure. 12+ years designing explainable AI systems for utilities, grid, and edge environments.
Nida Sahar Rafee
Co-Founder & COO
Cloud Infrastructure · Edge Networks
Led IT infrastructure at scale for telco and edge networks. High-availability systems design for large-scale distributed infrastructure deployments.
Devi Prasad Vuriti
Power Domain Lead
Power Generation · Transmission · DISCOM Operations
20+ years across generation, transmission, energy auditing, WHR and solar integration. Predictive reliability and governance dashboards for industrial leadership.
Kameswara Rao Tangudu
Integration & Solutioning Lead
OT/IT Convergence · API Governance · Utilities
Enterprise integration across mobility, utilities and safety domains. Middleware platforms and interoperability frameworks connecting OT/IT systems across regulated infrastructure environments.
Track Record
The founding team brings direct hands-on experience from power sector AI deployments in prior roles — including shadow-mode pilots covering outage management and compliance traceability at large-scale transmission and generation infrastructure. This operational experience is the architectural foundation of the SkyEdgeAI platform. Both founders are committed full-time.
Company
SkyEdgeAI — Purpose-Built for Infrastructure Governance
Founded 2025 · Bengaluru, Karnataka
Origin
SkyEdgeAI was purpose-built to solve a gap we observed repeatedly: infrastructure operators in regulated sectors have world-class operational systems but no equivalent governance intelligence layer.
Operational systems run the grid. Nothing was synthesising governance evidence from it.
We built the first read-only governance overlay purpose-designed for audit scrutiny — not operational control.
Operational systems run the grid. Nothing was synthesising governance evidence from it.
We built the first read-only governance overlay purpose-designed for audit scrutiny — not operational control.
This Grant
Validate TRL 4 → 6 for the DISCOM governance platform — demonstrating shadow-mode audit readiness in a Karnataka DISCOM environment, with STPI Bengaluru as institutional partner.
What We Do
- Detect cross-system risk patterns before they become failures — giving DISCOM operational teams the evidence basis to act hours earlier than any existing process allows
- Make every operational decision institutionally defensible — the field dispatch, the revenue protection intervention, the maintenance work order all carry a documented, tamper-evident justification basis
- Generate the continuous governance record that transforms a DISCOM from an organisation that explains failures after they happen into one that demonstrates proactive, evidenced governance to KERC, CAG, and the Energy Department
Who We Serve
- Primary: State power distribution utilities (DISCOMs / ESCOMs) across India
- Audit & Compliance teams who need continuous governance visibility
- Adjacent: Water utilities, gas distribution, transmission companies — same governance gap
The Problem
A Governance Gap That Regulators Are Already Documenting
DISCOMs generate vast operational data — but governance of that data remains reactive, fragmented, and audit-vulnerable.
The Root Cause — Cross-System Blindness
No system sees the intersection — where a correlated pattern across SCADA, OMS, MDMS, and ERP simultaneously signals a governance risk forming. Each system sees its own data. None sees what the combination means. By the time a manual review cycle catches the pattern, the failure has already occurred, the consumer complaint has been filed, and the KERC finding is forming.
The Compounding Problem — Accountability Paralysis
In a government DISCOM, acting without a documented cross-system evidence basis carries as much accountability risk as not acting at all. A distribution engineer who senses something is wrong faces a rational choice: act on intuition and risk accountability exposure — or wait for certainty, by which point the failure has occurred. This paralysis is systemic, not individual. It is the institutional consequence of having no continuous cross-system governance intelligence layer.
Consequences
- Late risk identification — systemic issues surface weeks after occurrence
- Manual compliance evidence — fragmented, inconsistent, gap-dependent
- Audit preparation takes 3–6 months per compliance cycle
- Regulatory penalty exposure — Rs. 1–10 Cr per DISCOM per SERC cycle
Why Karnataka, Why Now
- KERC is among India's most active regulatory commissions — BESCOM's AT&C losses remain above 11% against a national best-practice benchmark of under 8% (KERC Tariff Order FY24)
- CAG audits increasingly demand digital audit trails — manual reconstruction is no longer sufficient
- Data volume from SCADA, OMS, MDMS, AMR systems has outpaced manual governance capacity
- Karnataka's RAPDRP-established IT infrastructure is the most DISCOM-digitised in India — the integration foundation already exists
Primary Impact
Every person in the DISCOM accountability chain — from the MD who faces KERC and CAG scrutiny, to the distribution engineer who needs a defensible basis to act early, to the Audit team who currently spend months reconstructing evidence that this system generates continuously.
The governance infrastructure that makes the entire accountability chain functional.
The governance infrastructure that makes the entire accountability chain functional.
Market Timing
2026: A Tipping Point for DISCOM Governance
Four forces are converging — and they are converging fastest in Karnataka.
01
Regulatory Pressure Intensifying
KERC and CERC compliance timelines are tightening. CAG audit standards are rising. AT&C loss reduction targets create operational urgency — and regulators are demanding evidence of continuous monitoring, not annual snapshots.
02
Data Volume Has Outpaced Manual Governance
SCADA, OMS, MDMS, and AMR systems are generating unprecedented operational data volumes. Human review cycles cannot keep pace. Governance must become automated to remain meaningful.
03
Manual Processes Are Visibly Failing
Audit preparation takes months. Risk identification lags by weeks. Compliance evidence is fragmented and inconsistent across cycles. This is not a technology gap — it is a governance architecture gap.
04
Karnataka Is the Right Starting Point
Karnataka has India's most digitised DISCOM infrastructure — RAPDRP-established across 64 towns. BESCOM serves 30 million+ consumers. KERC is the most active state regulator. The foundation for AI governance exists here first.
The window for AI-enabled governance is open now. Karnataka is the right place to validate it.
Competitive Landscape
No One Else Does What We Do
Every existing solution either runs the grid, controls it, or audits it manually — once a year. No solution makes early action institutionally defensible for government officers or transforms the MD's accountability position.
| Solution | Read-Only? | AI-Native? | Explainable AI? | Audit Evidence? | Domain-Specific? | Defensible at Decision Point? | Transforms MD Accountability? |
|---|---|---|---|---|---|---|---|
| SkyEdgeAI | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |
| SAP / Oracle ERP | ✗ | ✗ | ✗ | ✗ | ~ Generic | ✗ | ✗ |
| GE / Siemens SCADA | ✗ | ~ | ✗ | ✗ | ~ Control only | ✗ | ✗ |
| Big 4 Audit Firms | ✓ | ✗ | ~ | ~ | ~ Point-in-time | ✗ | ✗ |
| Generic GRC Tools | ~ | ✗ | ~ | ~ | ✗ | ✗ | ✗ |
| TCS / Infosys SI | ✗ | ~ | ✗ | ~ | ✗ | ✗ | ✗ |
What "Explainable AI ✓" Means in Practice
For every anomaly detected, SkyEdgeAI generates a plain-language compliance narrative identifying the originating data source, the detection rationale, and the specific KERC/CERC obligation implicated. An auditor reads a document — not a model score requiring interpretation.
Gap Analysis
Why Existing Tools Cannot See It — and What Changes When They Can
Each tool was designed for one system. The intersection is invisible by design. Solving that invisibility changes three things simultaneously — at the operational, organisational, and institutional level.
Why Existing Tools Cannot See the Risk
The Cross-System Blindness
Each system sees one dimension of the problem:
- SCADA sees voltage deviations — normal equipment noise
- OMS sees two brief outages — within tolerance
- MDMS sees an AT&C spike — possibly seasonal
- ERP sees no maintenance work orders — nothing flagged
The intersection of all four simultaneously in the same zone, in the same 72-hour window, is not noise — it is the early signature of a technical or commercial governance failure. No single system sees it. No manual review cycle catches it in time.
By the Time Existing Approaches See It
- The outage cascade has occurred — consumers are affected
- The consumer complaint has been filed — KERC CGRF is engaged
- The AT&C loss has accumulated — revenue is unrecoverable
- The KERC finding is forming — penalty exposure is live
What Changes When the Intersection Becomes Visible
Operational — Early Action Becomes Possible
A cross-system signal surfaces hours before any single-system review would catch it. Distribution engineers can dispatch field teams, direct revenue protection resources, and raise maintenance work orders before the failure occurs — not after the cascade, the complaint, and the penalty.
Organisational — Early Action Becomes Defensible
A GuardianLedger™ event is the documented, tamper-evident basis for every early decision. The engineer who acts on it is institutionally protected. The paralysis is resolved not by changing people — but by giving them the evidence basis that makes early action safe.
Institutional — The Accountability Position Transforms
The DISCOM moves from explaining failures after they happen to demonstrating proactive, evidenced governance to KERC, CAG, and the Energy Department. The MD's relationship with every accountability body above them changes permanently — from adversarial examination to demonstrated governance.
The Solution
A Read-Only Governance Overlay — Not a Grid Control System
Before the architecture, before the steps, before the KPIs — this is what SkyEdgeAI is and what it is not. Everything else follows from this.
What SkyEdgeAI Is
A read-only governance overlay that sits above existing operational systems — SCADA, OMS, MDMS, ERP — reads them in parallel without touching their operations, detects cross-system risk patterns before they become failures, makes every operational response institutionally defensible, and creates the continuous governance record that transforms how a DISCOM relates to its regulators.
It is not a control system. It is not a replacement. It is the governance intelligence layer that existing systems were never designed to provide.
It is not a control system. It is not a replacement. It is the governance intelligence layer that existing systems were never designed to provide.
What It Is Not
- Not an operational system — it does not run the grid or manage outages
- Not a reporting tool — evidence is generated continuously, not assembled before an audit
- Not a replacement — works across existing SCADA, OMS, MDMS, ERP with no migration
- Not a decision-maker — the platform advises; human accountability is retained at every point
Design Guarantees
🔒
No Operational Authority
The system advises. It never commands. No output touches grid control.
👁️
Read-Only Integration
Connects to SCADA, OMS, MDMS, ERP via read-only interfaces. No write access. Ever.
🧑⚖️
Human Accountability Retained
Every governance event is attributable to a human decision-maker. AI advises; humans decide.
🏛️
No System Replacement
Works across existing SCADA, OMS, MDMS, ERP — no migration, no disruption.
Technical Architecture
Read-Only by Design — How the System Works
Data Flow
SCADA
Event logs
Event logs
OMS
Outages
Outages
MDMS
Losses
Losses
ERP
Billing
Billing
↓
Read-Only Ingestion Layer
Log forwarders / read-only API connectors via DMZ · No agent on OT network · All reads logged
Log forwarders / read-only API connectors via DMZ · No agent on OT network · All reads logged
↓
Proprietary AI Engine
Anomaly detection & pattern correlation · Domain-specific · Patent-protected · No open-source components
Anomaly detection & pattern correlation · Domain-specific · Patent-protected · No open-source components
↓
GuardianLedger™
Immutable evidence store
Immutable evidence store
Compliance Index
KERC/CERC mapped
KERC/CERC mapped
Integration Scope
- SCADA: Hansen, Oracle, GE, or vendor-neutral event log adapters
- OMS: Standard outage event APIs
- MDMS: AMR/AMI meter data feeds
- ERP: Oracle, SAP, or custom billing adapters
OT Network Safety
Access uses network-isolated log forwarders or scheduled batch exports from read-only API endpoints through a DMZ connection. No agent is installed on operational technology infrastructure. Every data read is timestamped and logged with user and scope.
Deployment
On-premise at the DISCOM's site. No cloud dependency. All data stays within the DISCOM's own infrastructure boundary.
System Logic
Five Steps. Nothing Written Back. No Command Goes Out.
From raw operational data to defensible operational action to audit-ready evidence — the platform reads and advises; the human decides and acts; GuardianLedger™ documents both.
01
Read-Only Ingestion
DISCOM data — historical datasets and mirrored real-time extracts from SCADA, OMS, MDMS, and ERP — is ingested through isolated, fully logged read-only connectors. No write access. No operational dependency.
02
Cross-System Risk Pattern Detection
The proprietary AI engine correlates signals across all four systems simultaneously — detecting the intersection pattern that no single system sees. A voltage deviation cluster in SCADA, two OMS outages, an MDMS AT&C spike in the same zone, in the same 72-hour window: that intersection is the risk signal. Purpose-built, patent-protected, no open-source components.
03
Actionable, Institutionally Defensible Signal
GuardianLedger™ generates a plain-language advisory — the cross-system evidence, the KERC/CERC obligation implicated, and the recommended operational response — delivered to the operational team before the failure occurs. This is not a model score. It is a documented, tamper-evident basis for a government decision. The field team dispatch, the revenue protection intervention, the maintenance work order — all carry this reference as their institutional justification.
04
The Operational Team Acts — With Documented Justification
The human decision-maker acts on the advisory signal. The platform never acts. Human accountability is retained at all times. The decision — and its evidence basis — is recorded before the action is taken.
05
GuardianLedger™ Documents Both Detection and Response
The continuous governance record captures the risk signal, the KERC/CERC obligation mapping, the advisory output, and the operational response — timestamped, cryptographically chained, tamper-evident. When KERC asks what governance processes were in place, the answer is not a description. It is a document.
What Changes for the People in the Accountability Chain
MD
Demonstrated governance replaces adversarial examination — at KERC, CAG, and Energy Department
Distribution Engineer
Acts early on a GuardianLedger™ signal — institutionally protected, not exposed
Audit & Compliance
Evidence generated continuously — not assembled in the months before an audit
CFO
Prevented cascade failures and RDSS eligibility — operational cost avoidance, not licence ROI
Deeptech & IP
Proprietary Technology at a Genuine Research Frontier
The output is a compliance document. The engineering sits at the intersection of five disciplines — none of which alone reaches the frontier. The problem has remained unsolved not for lack of trying, but because the intersection is the hard part.
OT/IT
Integration
Integration
Domain-Specific
AI/ML
AI/ML
Cryptographic
Evidence
Evidence
Regulatory
Knowledge
Knowledge
Explainability
Engineering
Engineering
GuardianLedger™
Immutability Mechanism — Cryptographic Evidence Domain
Implemented via a cryptographically hashed append-only event store — each governance event is assigned a SHA-256 hash chained to the prior entry. Tampering is cryptographically detectable. Performant on-premise within DISCOM compute constraints — no blockchain overhead required.
Five Properties
- Event immutability — once recorded, governance events cannot be altered; any tampering is detectable
- Cross-system correlation — unifies events from SCADA, OMS, and MDMS into a single auditable timeline
- Regulatory obligation mapping — each event tagged to specific KERC/CERC requirements; new orders ingested on issuance
- Explainable evidence output — plain-language compliance narratives; auditors read documents, not model scores
- Advisory-only architecture — every entry attributable to a human decision-maker
Proprietary AI Engine
Why Each Discipline Alone Fails — The Intersection Argument
- AI/ML alone cannot meet the OT deployment constraints or produce institutionally defensible output from regulatory knowledge
- OT/IT integration alone can get the data but cannot synthesise governance evidence from it
- Cryptographic architecture alone can create tamper-evident records but cannot determine what is worth recording or why
- Regulatory knowledge engineering alone cannot detect which operational events implicate which obligations in real time
- Explainability engineering alone cannot produce institutionally defensible output for a government accountability context
The intersection is the research frontier. That is why this has not been solved before.
IP — Three Layers · Technical Uncertainties the Pilot Resolves
IP Layers
- Filed patents — GuardianLedger™ + anomaly detection methodology
- Training corpus — cannot be replicated without domain access
- Regulatory knowledge graph — years of KERC/CERC filings encoded
TRL Uncertainties (Pilot Resolves)
- Model generalisability across DISCOM environments
- KERC evidentiary acceptance of GuardianLedger™ artifacts
- Live production false positive rate vs. calibration data
- Institutional adoption behaviour under the accountability framework
Technology Readiness
TRL 4 → TRL 6: A Validated Core to a Demonstrated Environment
The grant does not fund a concept — it validates a working system in a real DISCOM environment. TRL 4 to 6 sits squarely within the ELEVATE NXT programme scope of TRL 3–8.
1
Basic
2
Concept
3
Proof of Concept
4
Current — Core Validated
5
Relevant Env.
6
Target — Demonstrated
7+
Scale
TRL 4 — Where We Are Now
Core components validated in controlled environments. Evidenced by the founding team's prior hands-on experience deploying shadow-mode AI pilots covering outage management and compliance traceability at large-scale transmission and generation infrastructure — the same architectural patterns and integration mechanisms underpin the SkyEdgeAI platform.
TRL 6 — Grant Target
Platform demonstrated in a relevant Karnataka DISCOM environment through a structured shadow-mode pilot — producing a validated KPI dossier, compliance traceability evidence, and stakeholder sign-off from the DISCOM Audit & Compliance team.
During the Grant
Analytics hardening for DISCOM data variability
Expanded compliance traceability framework (KERC/CERC obligations)
Stakeholder validation with DISCOM Audit & Compliance teams
Pilot Design
Shadow-Mode Validation — Zero Risk to Operations
One Karnataka DISCOM division. Read-only. Advisory. The DISCOM takes on no operational or data risk.
Pilot Scope
- Site: One Karnataka ESCOM division (primary pilot site)
- Mode: Read-only, advisory — no operational integration
- Data: Historical datasets + mirrored real-time extracts
- Duration: 6 months
Pilot Activities
- Read-only data ingestion from SCADA, OMS, MDMS, ERP via isolated connectors
- Risk anomaly detection using proprietary AI engine
- Compliance mapping against KERC/CERC regulatory obligation taxonomy
- GuardianLedger™ evidence generation and audit artifact production
- Structured validation sessions with DISCOM Audit & Compliance teams
Explicit Exclusions
- No grid control or operational commands
- No billing or consumer operations
- No procurement implications
- No data leaves the DISCOM's infrastructure boundary
DISCOM Engagement Pathway
Securing a Karnataka ESCOM as pilot site is Milestone 1 of the programme — a known, managed dependency with a clear mechanism, not an assumption.
- STPI Bengaluru provides the institutional channel to reach ESCOM MD level — bypassing standard vendor queues; engagement begins on grant day one
- Data access agreement sits below the KTPP tender threshold — no competitive tender required; approval rests with the ESCOM MD with involvement from IT and Audit heads
- The read-only, on-premise, grant-funded design means the ESCOM takes on near-zero legal, operational, and financial risk — the lowest-friction entry point possible
- Realistic commitment: 90–120 days to signed data access agreement, accounting for government legal review and committee scheduling
Data Ownership
The DISCOM retains full ownership and control of all data at all times. Access is read-only and fully logged. Every data read is timestamped and attributable. Even if SkyEdgeAI were acquired, the DISCOM has no obligation and retains all data.
Measurable Outcomes
KPIs Measured Against a Documented Pre-Pilot Baseline
Baseline documentation is the first activity — we establish current-state metrics before any SkyEdgeAI outputs are active.
| KPI | Current Baseline (Est.) | Target Post-Pilot |
|---|---|---|
| Risk detection coverage | Reactive / post-event only | >80% of systemic risks identified proactively |
| Time to risk identification | 2–4 weeks (manual review cycle) | <4 hours (automated detection) |
| Compliance traceability index | Fragmented, manual, <20% automated | 80–90% automated linkage to KERC/CERC |
| Audit preparation effort | 3–6 months per compliance cycle | <4 weeks (40–60% reduction) |
| Evidence completeness score | Inconsistent; gap-dependent | 95%+ from Day 1 of pilot |
Quality KPIs
False Positive Rate
Target: <15% of flagged events. Advisory-only architecture means false positives generate no operational impact — measured and reduced iteratively.
Stakeholder Acceptance
Structured survey with DISCOM Audit & Compliance teams at midpoint and close. Acceptance score is a formal KPI — not a qualitative observation.
Operational KPIs — Measuring Behavioural Change, Not Just Evidence Output
| Operational KPI | Current Baseline | Target Post-Pilot |
|---|---|---|
| Mean time: risk signal → operational response Measures whether the DISCOM is actually acting differently |
Days to weeks (manual review → escalation → justification cycle) | <4 hours from GuardianLedger™ event to documented response |
| Proactive work orders triggered by system signal Measures decision-making paralysis conversion |
Zero — all current work orders are reactive or pre-scheduled | Defined monthly target per division, documented with evidence reference |
| Revenue protection interventions: signal-directed vs. blanket Measures precision improvement in AT&C recovery |
100% blanket zone audits — undirected, slow, politically sensitive | Defined % of interventions directed by cross-system feeder-level signal |
Go-To-Market
Three Parallel Tracks. One Reference. National Moat.
Government DISCOM procurement is 24–30 months from first engagement to signed licence. We have designed the company around that cycle — not against it.
Track A · Pilot
Grant Period — Months 1–12
Below KTPP threshold. No tender required.
- M1–3: STPI Bengaluru institutional engagement → BESCOM MD
- M3–4: Data access agreement signed (90–120 days — government legal review + committee scheduling)
- M4–10: 6-month shadow-mode pilot, one BESCOM division
- M10–12: KPI dossier, stakeholder validation, security audit certificate
Output
Validated platform + KPI evidence + Audit & Compliance team endorsement + security clearance
Track B · Commercial
Months 12–30 · KTPP/GeM Process
- M12–18: BESCOM Board memo — MD + IT + Audit heads. Board of Directors approval (3–6 months at a financially-stressed ESCOM)
- GeM: SkyEdgeAI is a registered GeM vendor — custom bid via GeM eliminates full KTPP open tender, compressing procurement by 4–6 months
- M18–26: GeM custom bid published, technical evaluation (pilot dossier = sole validated evidence), L1 selection
- M26–30: Contract signed, performance security, first Rs. 50L–1 Cr annual licence payment
Why SkyEdgeAI Wins the GeM Bid
No other vendor will have run a shadow-mode proof-of-concept at BESCOM. The tender specification is informed by our pilot learnings — first-mover advantage translates directly into specification ownership.
Track C · Policy
Months 6–36 · Runs in Parallel
- M6–12: Energy Department informal briefings via STPI — begins during pilot, not after
- M12–18: RDSS AI/ML use case registration (REC/PFC portal — active call) — converts licence from ESCOM opex to central grant-eligible spend
- M18–36: Karnataka Energy Department G.O. issued — recommends all five ESCOMs evaluate validated platform
- M24–36: KERC accepts GuardianLedger™ artifacts at BESCOM annual performance review — establishes regulatory evidence standard
The Moat This Creates
Each subsequent ESCOM inherits BESCOM's credibility — no repeat pilot required. KERC acceptance + G.O. + RDSS registration means Year 3–6 cycles are 50–60% shorter than Year 1–2.
Buyer Motivation by Persona
- MD: Accountability protection — demonstrated governance vs. adversarial examination
- CFO: Prevented cascade failures + RDSS eligibility — operational cost avoidance, not licence ROI
- Audit Head: Workload transformation — continuous evidence vs. months of manual assembly
- CIO: IT governance documentation + personal CAG audit protection
- Energy Dept: Standardised continuous governance across all 5 ESCOMs — institutional champion
Honest Revenue Trajectory
First licence: Month 26–30. Karnataka ARR at Year 4: Rs. 1–2 Cr. 5–8 DISCOMs, Rs. 3.5–8 Cr ARR at Year 5–6.
Why the Slow Cycle Is the Moat
Any competitor entering post-BESCOM faces the same 24-month cycle — without pilot evidence, KERC acceptance, G.O., or RDSS registration. First-mover in government time = 3–4 years of defensible exclusivity.
ROI & Value Case
A Rs. 1 Crore Grant to Validate a Rs. 100–200 Crore Annual Solution
Every figure sourced from KERC Tariff Orders, PFC Annual Reports, UDAY scheme data, and CARE Ratings — not estimated.
BESCOM Baseline (FY24)
Rs. 34,000 Cr
Annual Revenue
Source: BESCOM FY24 filings
~12%
AT&C Loss
vs. <8% best practice (KERC FY24)
30M+
Consumers
Source: CARE Ratings July 2024
Annual Value per DISCOM
| ROI Category | Conservative | Realistic | Source |
|---|---|---|---|
| AT&C governance contribution 0.3–0.5% improvement in commercial losses |
Rs. 100 Cr | Rs. 200 Cr | PFC / UDAY data |
| Audit preparation saving 40–60% reduction in person-hours |
Rs. 2 Cr | Rs. 7 Cr | PFC benchmarks |
| Regulatory penalty avoidance 30–50% of KERC penalty exposure |
Rs. 2 Cr | Rs. 5 Cr | KERC/CERC orders |
| RDSS scoring uplift Protects central grant eligibility |
Rs. 50–500 Cr in RDSS grants protected | MoP RDSS guidelines | |
| Prevented cascade failures Emergency vs. planned maintenance differential |
Rs. 2–3 Cr | Rs. 5–10 Cr | Equipment cost differential; DISCOM maintenance records |
| Revenue protection precision Targeted vs. blanket zone interventions |
Faster AT&C recovery, lower cost per rupee recovered — directional; quantified post-pilot | Operational baseline established during pilot | |
The Economics
100x – 400x
Return on licence cost
Rs. 50L–1 Cr licence · Rs. 104–212 Cr annual value
Rs. 50L–1 Cr licence · Rs. 104–212 Cr annual value
The Grant ROI
Rs. 1 Cr grant to validate a solution that delivers Rs. 100–200 Cr of annual value at BESCOM alone — recurring every year the platform operates.
Why RDSS Changes the CFO Calculation
Once registered as an RDSS AI/ML use case (active REC/PFC call, May 2026), DISCOMs can fund the licence from central government scheme money — not their own distressed operating budget. The licence cost moves from an ESCOM opex line to a grant-eligible investment.
BESCOM Net Worth Context
BESCOM carries negative net worth of Rs. 6,442 Cr (KERC APR April 2026). The cost of not having continuous governance visibility — KERC penalties, RDSS de-eligibility, CAG audit findings — is higher than the licence. That is the CFO argument.
The Institutional Value — Why the MD Authorises It
The financial return is why the CFO approves it. The institutional value is why the MD authorises it: protection of the entire accountability chain — from MD to distribution engineer — through documented, defensible, continuous governance evidence. A DISCOM MD who can demonstrate proactive governance to KERC, CAG, and the Energy Department is not just managing a compliance tool. They are changing their institutional relationship with every accountability body above them.
Market Opportunity
A Rs. 500–1,000 Crore Governance Tooling Opportunity in India
Sized on per-DISCOM governance licence pricing — not total IT spend.
Market Sizing
| Segment | Basis | Size |
|---|---|---|
| TAM | ~100 DISCOMs × Rs. 50L–1Cr governance licence/yr | Rs. 500–1,000 Cr |
| SAM | ~50 digitally mature DISCOMs with SCADA/OMS/MDMS stack and active regulatory scrutiny | Rs. 250–500 Cr |
| SOM Year 2.5–4 |
5 Karnataka ESCOMs × Rs. 50L–1Cr/yr (GeM procurement → G.O. → sequential adoption) | Rs. 2.5–5 Cr |
Market sizing anchored to governance tooling licence pricing — not total IT spend. Governance is a dedicated, growing budget line as regulatory pressure intensifies.
Adjacent Expansion
Beyond DISCOMs: water utilities, gas distribution networks, and transmission companies (e.g., KPTCL) share the same governance gap profile. DISCOM validation creates a replicable template across regulated infrastructure.
Karnataka Tailwind
- Karnataka has India's most digitised DISCOM infrastructure — RAPDRP across 64 towns; the integration foundation exists here first
- BESCOM alone serves 30 million+ consumers — governance improvement here has statewide multiplier effects
- Karnataka's clean energy push (14.6 GW installed renewable capacity) adds grid governance complexity and urgency
- KERC is among India's most active state regulatory commissions — highest scrutiny, highest value from continuous monitoring
Why Governance Tooling Is Growing
- KERC/CERC compliance timelines tightening; AT&C targets create urgency for continuous monitoring
- CAG audits increasingly demand digital audit trails — manual reconstruction no longer sufficient
- India's 24×7 Power for All commitments and renewable integration adding grid complexity governance cannot track manually
Socio-Economic Impact
Governance Failures Cost Karnataka More Than You Think
Current Cost of Poor DISCOM Governance
| Problem | Data Point | Source |
|---|---|---|
| AT&C losses (Karnataka) | BESCOM ~11.5%; state avg higher vs. best practice <8% | KERC Tariff Order FY24; PFC Annual Report 2022-23 |
| AT&C financial impact | Each 1% loss = ~Rs. 200–400 Cr/yr for a major DISCOM | UDAY scheme data; PFC state utility reports |
| Manual audit prep cost | Rs. 5–15 Cr/yr per DISCOM in person-hours | PFC Performance Report; operator estimates |
| Time to identify risk | 2–4 weeks (reactive manual review) | DISCOM audit cycles; operational baseline |
| Regulatory penalty exposure | Rs. 1–10 Cr/DISCOM/SERC cycle | KERC/CERC public penalty orders |
How SkyEdgeAI Reduces This Cost
- Audit preparation: 40–60% effort reduction via automated evidence generation
- Time to risk identification: 2–4 weeks → <4 hours (proprietary anomaly detection)
- Compliance traceability: 80–90% automated linkage to KERC/CERC obligations
- Evidence completeness: 95%+ from Day 1 (GuardianLedger™ continuous capture)
- Consumer grievance patterns: Early warning reduces escalation — anomaly detection flags deteriorating service metrics
Institutional Socio-Economic Impact
- Public accountability transformed: A DISCOM with continuous, evidenced governance is accountable in a fundamentally different way — consumers have stronger rights, better redressal, and a utility whose governance record is transparent and auditable
- National governance standard: Karnataka demonstrating continuous, standardised governance evidence sets the reference for how government utilities in India relate to their regulators — a governance legacy beyond technology deployment
- Employment & skill creation: AI/ML/data science roles in Karnataka; skill uplift for DISCOM operational and compliance teams; replicable governance framework as national knowledge infrastructure
Budget & Grant Utilisation
Rs. 1 Crore — Every Rupee Scoped to TRL Validation
No spend on sales, marketing, production deployment, or live operations. One DISCOM division. Fully costed.
Line-Item Budget
| # | Line Item | Amount |
|---|---|---|
| 1 | AI model training & DISCOM data variability hardening | Rs. 22L |
| 2 | Compliance traceability framework development | Rs. 18L |
| 3 | Shadow-mode pilot operations (1 DISCOM division, 6 months) | Rs. 25L |
| 4 | Security audit, pen testing & governance review | Rs. 12L |
| 5 | KPI validation dossier & grant reporting | Rs. 10L |
| 6 | Team & project management (grant period) | Rs. 8L |
| 7 | Contingency (10%) | Rs. 5L |
| Total | Rs. 1 Cr | |
Tranche-Linked Milestones
Tranche 1
Rs. 25L
Grant acceptance + team onboarding + Karnataka ESCOM data access agreement signed (90–120 days via STPI facilitation — below KTPP tender threshold, MD-level approval)
Tranche 2
Rs. 30L
AI model hardening complete + GuardianLedger™ v1 deployed in shadow mode at pilot site
Tranche 3
Rs. 25L
Compliance framework validated with DISCOM Audit team + KPI baseline documented
Tranche 4
Rs. 20L
Shadow-mode pilot complete + KPI validation dossier submitted + stakeholder sign-off received
Karnataka Relevance
Karnataka as the Right First Environment for This Technology
Not just the most convenient — the most consequential. What works here sets the national standard.
Why Karnataka Specifically
- India's most digitised DISCOM infrastructure — RAPDRP across 64 towns; the integration foundation for AI governance exists here first
- KERC is India's most active state regulatory commission — governance tooling is most meaningful where scrutiny is highest
- STPI Bengaluru provides institutional credibility and formal ESCOM engagement pathway
- ELEVATE NXT ecosystem provides post-grant pathways for reference customer development and G.O. adoption
What This Grant Delivers for Karnataka
- Addresses increasing audit and governance pressure in Karnataka ESCOMs with a validated technology response
- Establishes a Karnataka-first, India-first reference for AI-enabled infrastructure governance
- Creates reusable governance frameworks and KPI benchmarks that become public infrastructure for the sector
- Positions Karnataka as the governance technology leader in India's power distribution sector
- Karnataka's 14.6 GW renewable energy integration adds grid governance complexity — intermittency creates new cross-system anomaly categories that manual review cannot track; this platform handles them automatically
Non-Binding Safeguards
- No procurement implication from this grant — future adoption subject to standard G.O. processes
- No vendor lock-in — DISCOM retains all data and all rights
- Even a non-adoption outcome delivers validated workflows and benchmarks that benefit Karnataka's regulatory ecosystem
- If Karnataka ESCOMs demonstrate continuous, evidenced governance, the national conversation about utility accountability changes — that is a governance legacy Karnataka sets for India
BESCOM Scale
30M+
Consumers served by BESCOM alone
14.6GW
Karnataka installed renewable capacity
64
Towns with RAPDRP-established IT infrastructure
The Ask
A Safe, Transparent Way for Karnataka to Validate What Comes Next
Scope Lock
- Read-only, advisory system
- No operational authority
- No system replacement
Risk Posture
- No operational risk — shadow mode only
- No liability transfer — DISCOM retains full data control
- Financial risk limited to milestone-linked tranches
Failure Is Acceptable
- Even if some KPIs fall short, the pilot delivers validated governance workflows, auditable artifacts, KPI benchmarks, and documented learnings
- Both success and partial success have value for Karnataka's regulatory infrastructure
Rs. 1 Crore
Grant to validate TRL 4 → 6 for India's first read-only governance overlay for DISCOM audit readiness.
A safe, transparent way for Karnataka to validate next-generation infrastructure governance — before future policy decisions are made.
A safe, transparent way for Karnataka to validate next-generation infrastructure governance — before future policy decisions are made.
Post-Grant Trajectory
First BESCOM licence signed at Month 26–30 via GeM procurement. Karnataka G.O. and RDSS AI/ML use case registration follow during Years 2–3, creating the policy infrastructure for sequential ESCOM adoption. 5–8 DISCOMs, Rs. 3.5–8 Cr ARR by Year 5–6 — patient capital, defensible moat. The KPI validation dossier anchors SkyEdgeAI's seed/pre-Series A raise targeting energy governance and infrastructure AI investors.