Depth bar: Due-diligence dossier for Stamped Energy outreach. Estimates marked
[~]. Directory phones[dir]. Never invent bill numbers, court outcomes, or emails.
1. Company overview & snapshot
1.1 Legal identity & corporate structure
Amba Shakti Steels Limited — CIN U27104DL1989PLC127527, incorporated 16 Mar 1989, ROC Delhi; formerly Amba Steels (unit of K.P. Steel Products) context in rating notes. Directors include Pankaj Goel (DIN 00224355), Manik Goel (DIN 08397124), and others. Group spans steel, LPG cylinders, real estate, biofuels entities — confirm contracting entity = ASSL Muzaffarnagar plant.
1.2 What they make & where money comes from
TMT bars (Amba Shakti / Tricore brands), MS billets. Rating note (Jan 2024) cited installed 360,000 MTPA (TMT 180k + billets 180k) with grades Fe415–Fe550D. Lead revenue ~₹642 Cr FY25. Direct hot charging marketed. Credit rating BBB-/stable class historically (Infomerics note) — directional only.
1.3 Plants, addresses & footprint
Pilot: 7 Km Stone, Meerut Road, Muzaffarnagar 251003. Phone +91 9837084505; info@ambashaktisteels.com. MoRTH NOC context for 132 kV UG XLPE cables along NH-58 — signals high electrical demand growth. Group also lists other UP/MP plants — anchor ASSL Meerut Road HT first.
1.4 Leadership & CRM map
Primary kit: Manik Goel, CEO Amba Shakti Tricore (green steel / renewable public narrative; personal LinkedIn slug not confirmed — use company page amba-shakti-official + info@). Secondary: Pankaj Goel (Vice-Chairman), S.K. Goel (Founder), plant Head Electrical (ask). Path: promoter → electrical → 90-day scope.
1.5 Recent news (24 months) & timing for Stamped
Public green-steel / renewable (solar+wind) manufacturing claims on LinkedIn company posts; 132 kV cable laying proposals; expansion narrative to higher melt capacity in lead research. Timing: challenge MD after hot-charge + RE claims — “green process, bill still leaks.”
2. Energy profile
DISCOM / supply (name early): Paschimanchal (PVVNL) / UPPCL family (KESCO, MVVNL, PuVVNL, DVVNL).
2.1 Bill band, tariff & demand
Large integrated plant: hypothesis ₹80 lakh–₹3 Cr+/month [~]. Investigate 132 kV / CMD step-up. Verify invoice ≥ ₹30L easily or kill.
2.2 Generation, fuel & renewables
Company claims elimination of coal/FO and advanced pollution control; captive solar and wind claimed in LinkedIn posts for Tricore manufacturing. Treat as marketing until PPA/kWp verified. Hot charging reduces reheat fuel — electrical MD remains.
2.3 EnMS, PAT, ISO, BRSR
Private — no BRSR. ISO 50001 not evidenced. Green narrative ≠ EnMS maturity.
2.4 Likely ₹ leak categories (hypothesis)
Multi-IF campaign coincidence; mill start vs melt; RE curtailment vs grid MD; dust extraction idling; PF; 132 kV transformer losses/idle auxiliaries.
3.1 Process flow & critical loads
Build a one-page process map in discovery for Amba Shakti 7 Km Stone: scrap → multi-IF SMS → CCM with direct hot charging → high-speed rolling (Tricore TMT) → cooling → dispatch. Mark batch starts, heat-up/hold, changeovers, packing starts, utility headers and large motors. The key question is not simply “what consumes kWh?” but “which controllable operating event produces cost at the demand, tariff, PF or baseline level?” Capture production constraints before recommending sequencing. Quality, safety and customer delivery windows constrain what can move into lower-cost ToD slots.
Critical loads to inventory: incomer HT, process feeders (SMS/reactors/wet-end), furnace/TFH/HVAC banks, compressors, cooling towers, pumps, finishing/packaging motors, and any captive or DG bus. Ask which loads share a maximum-demand ratchet and which have separate demand contracts. For multi-block campuses, map which buildings share the pilot meter boundary.
3.2 Shifts, seasonality, production pattern
Determine if the site is continuous, campaign-based, multi-shift or heavily seasonal. Record scheduled breaks, weekly shutdowns, maintenance windows, SKU/changeover frequency and dispatch-driven overtime. A valid bill comparison must normalise for production volume, weather-sensitive HVAC, production mix and commissioning/ramp-up. If throughput is unstable, set success metrics around avoidable peaks, idle hours and verified invoice components rather than gross kWh alone. Ask which month in the last year produced the highest billed MD and what co-started that day.
3.3 Automation, metering, SCADA/EMS/DCS
No plant-specific SCADA/EMS vendor is assumed unless independently sourced for this dossier. Start with a data-access inventory: main-meter interval data; feeder/sub-meter exports; DCS/PLC historian tags; boiler/TFH/HVAC logs; compressor/chiller status; production schedule; and two to six PVVNL bills. Path A is a read-only connection/export from existing systems. Path B is structured CSV or meter export plus a production-event log. Both paths must exclude PLC writes, recipe changes and unapproved remote control. Cyber and OT approvals should be recorded before data leaves the site.
3.4 Capex / tech projects affecting energy
Ask about new lines, automation, solar, open access, utility upgrades, VFDs, compressor replacement, furnace/chiller work and product launches in the last 24 months. These can create a false “before” baseline and may have warranties or safety constraints. Stamped should position post-capex work as operational verification: ensure the asset is scheduled and used to achieve the expected bill result, rather than claiming credit for capex it did not cause.
4. Stamped Energy fit analysis
4.1 ICP scorecard
| Gate | Assessment |
|---|---|
| Geography (North India industrial) | Pass |
| Vertical / process intensity | Pass (see §3) |
| Bill ≥ ₹30 lakh/month HT | estimated ₹80L–₹3 Cr+/mo [~] — unknown until invoice |
| Decision speed / plant authority | Unknown until discovery |
| Data maturity | Unknown–medium; Path A/B TBD |
The account passes to a 20-minute discovery call, not automatically to a full integration.
4.2 Fit score rationale
Kit fit score 9/10 reflects process electric intensity, plausible HT exposure and a reachable plant/owner path. Score deductions for disclosure gaps, multi-entity groups, listed-company procurement, integrity flags, or bill uncertainty are handled in §6. The initial call must explicitly qualify bill band, site authority, accessible data and one measurable problem before committing engineering effort.
4.3 Wedge (parser-critical)
The strongest wedge is: IF campaign / mill start coincidence that still sets Paschimanchal MD after hot-charging and renewable claims — assign ₹ owners and verify with evidence
4.4 Objections & competitors
Likely alternatives: internal electrical/continuous-improvement teams; existing EMS/SCADA analytics; ISO/energy consultant; equipment OEM; “we already have solar/open access”; and for listed names, corporate sustainability routing. Response: Stamped does not replace controls or sell capex; it prioritises operational prescriptions, attaches an owner and reconciles results to the PVVNL bill. For a sophisticated site, offer a small read-only proof with explicit security/data boundaries. For a private site, lead with a concrete invoice problem rather than AI, ESG or digital-transformation language. If an EMS incumbent is present, emphasise read-only layer + WhatsApp assignment + invoice M&V.
4.5 Pilot design
Start with one meter boundary or one controllable utility/process cluster for 90 days (90-Day Bill Verification Program). Weeks 1–2: validate PVVNL bills, meter coverage, production context and baseline. Weeks 3–8: issue weekly ranked prescription cards covering MD sequencing, idle loads, thermal hold, air/HVAC, PF or tariff dispatch. Weeks 9–12: compare normalised bill components and production-adjusted indicators. Success: at least one action has an owner, measurable execution evidence and a defensible ₹/invoice movement. Kill criteria: bill below threshold, no data access, no plant owner, unstable operation without usable normalisation, or no controllable lever. Rollout path: additional feeders or sister units only after invoice proof.
5. Before you reach out
5.1 Discovery checklist
- Confirm the exact legal entity, site address, PVVNL account name, tariff and whether this is an HT/EHT bill.
- Verify monthly bill band in ₹ lakh/₹ Cr, sanctioned demand, billing demand and PF line.
- Identify the highest-load process, utility owner, shift pattern and current production constraint.
- Ask for the most recent MD event: when, what started, what it cost and who acted.
- Ask whether heaters/thermal systems, compressors, chillers, AHUs or pumps remain live during breaks/changeovers.
- Confirm existing EMS/SCADA/historian and what can be exported read-only.
- Confirm solar/captive/open-access context before discussing tariff-smart dispatch.
- Identify whether a site sponsor can approve a bounded 90-day proof and who must clear IT/OT.
- Request two PVVNL invoices, interval demand/energy data and a production/event calendar.
- Define a success measure tied to bill lines and production normalisation, not a generic dashboard KPI.
- Confirm champion tenure and email before first outreach send.
- Record any integrity/regulatory open items from §6.1 before proposing a commercial pilot fee.
5.2 Do not lead unwisely
- Do not lead with dashboards, AI buzzwords, generic kWh claims or a promised percentage reduction.
- Do not lead with ESG, carbon or ISO reporting first; lead with a controllable ₹ line on the PVVNL bill.
- Do not imply control-system writes, replacement of the EMS, compressor maintenance or solar EPC.
- Do not state the estimated bill band as fact; ask to verify it.
- Do not attach unrelated regional NGT/PCB matters to this company without named entity evidence.
- Do not challenge green claims adversarially; lead with MD despite RE.
5.3 Opening hooks (email / call / WhatsApp)
“At 7 Km Stone, IF and mill starts still set billed MD even with hot charging — we assign ₹ read-only.” Then ask which hurts more today—MD at starts, utility hold through changeovers, or off-shift auxiliaries. A valid next step is two bills plus the correct electrical POC, not a platform demo.
Extended operating hypotheses (call-prep depth)
Treat the following as checklists, not accusations. For each item, capture owner, constraint, last MD event date, and whether the load can move without quality loss. Stamped only acts where the plant owner accepts the prescription and the meter boundary can prove the rupee line on the next DISCOM invoice.
Demand-side sequencing. Map the top five motors or furnaces by kW rating and note whether their start windows overlap at shift change, after planned stops, after quality holds, or after utility recovery. Build a simple coincidence matrix for a typical production day. If two or more assets above roughly 20 percent of contract demand can start within the same 15-minute billing window, that window is a candidate for a first prescription card.
Thermal and hold energy. For furnaces, thermic-fluid heaters, dryers, chillers or air-handling units, ask what temperature or pressure is held during breaks, changeovers, waiting for quality clearance, or waiting for the next campaign. Idle hold often dominates avoidable kilowatt-hours even when operators believe the plant is already efficient. Distinguish safety-minimum hold from convenience hold and document who can authorise a change.
Compressed air and pumped loops. Inventory compressor count, pressure setpoints, unloaded run hours, and leak culture. For process pumps and cooling-water circuits, ask whether variable-frequency drives or on/off control exists and whether loops stay live for an entire weekend. These are common Path B wins when historian tags are missing but shift operators already know the waste pattern.
Tariff, power factor and open-access banking. On the invoice, separate energy, demand, power-factor incentive or penalty, and any open-access or solar adjustment lines. A renewable or open-access story can still leave maximum demand untouched. Banked units that expire or mismatch production timing are a separate commercial conversation from technical specific energy consumption.
Normalisation protocol for the 90-day program. Agree a production denominator (tonnes, batches, kilolitres processed, pairs, or operating hours), excluded shutdown days, weather adjustment for HVAC where relevant, and which invoice lines count as success. Without normalisation, a quieter month looks like a false win and destroys trust with technical buyers. Record the baseline window in writing before week-three prescriptions begin.
Security and OT boundary. Document the allowed interface (CSV export, historian read replica, supervised screen-share export), retention period, named data custodian, and that Stamped never writes PLC or DCS setpoints. Put this in the first scoping memo so IT or OT approval does not become a late surprise that burns a champion.
Commercial kill criteria (explicit). End the pilot early if: the bill gate fails; no named owner appears after week two; data is delayed more than two billing cycles; production mix changes so violently that normalisation is impossible; or compliance holds prevent any operational change. A clean kill protects the relationship better than a forced percentage claim.
Peer and cluster use. After invoice proof on one feeder, discuss sister feeders or nearby peer plants only as a second conversation. Do not sell a multi-site master agreement on first contact. For group companies, keep one CRM thread and one success story before expanding.
Finance and bill literacy. Ask who in accounts or works pays the HT bill, who disputes demand charges with the DISCOM, and whether anyone currently reconciles feeder estimates to the invoice. If nobody owns that reconciliation, Stamped’s bill-verification loop is itself the product, not a side feature of another dashboard.
People and shift reality. Identify which supervisors can actually stagger a start without waiting for board approval. A prescription without a shift owner dies in WhatsApp unread. Prefer two or three executable cards per week over a fifty-page audit that nobody implements.
6. Risks, flags, controversies & sources
6.1 Integrity / controversy / regulatory (search explicitly)
- PCB / historical: sector reports of past UPPCB fines for anti-pollution equipment non-use appear in aggregator synthesies for steel units — no 2024–2026 NGT judgment uniquely naming ASSL with a verified outcome located in July 2026 search. Search terms: Amba Shakti Steels Muzaffarnagar + (NGT|PCB|GST|lawsuit|fraud|labour).
- GST/lawsuit/fraud: none found establishing current adverse GST raid or fraud conviction.
- Promoter complexity: multi-entity Goel group — confirm decision unit and invoice entity.
- MSME payment delay listings on some aggregators: treat as credit-context noise, not outreach content.
Positive initiatives / energy posture (also searched): Direct hot charging; claimed captive solar/wind; APC systems; green steel/Tricore brand investment; 132 kV infrastructure intent.
6.2 Data quality flags
- The PVVNL service, bill band, sanctioned demand, tariff and site-meter boundary are research hypotheses unless supported by an actual invoice.
- Leadership titles, email patterns and LinkedIn availability must be confirmed immediately before outreach.
- Process, shift and automation details contain informed inference where public site disclosures are thin.
- Separate plant-specific fact from group-level ESG, revenue, certification and capacity disclosures.
- Confirm ASSL vs sister-plant HT account; verify RE capacity vs marketing.
6.3 Sources consulted
- ambashaktisteels.com about/contact; ambashaktigroup.com
- Infomerics rating PDF Jan 2024; MoRTH 132 kV NOC listing
- LinkedIn company amba-shakti-official posts (Manik Goel attribution)
- UP lead report; Tracxn CIN page
Extended call-prep narrative (Stamped field use)
B1. First 20 minutes of discovery
Open with process respect, not software. Confirm legal entity on the invoice, DISCOM name, and who owns the largest feeder MD event in the last 90 days. Ask for two invoices and one production calendar export before any platform demo. If the champion cannot produce bills within a week, treat the opportunity as stalled — Bill Verification Program without invoices is theatre. Record whether Path A (historian/meter export) or Path B (CSV) is realistic given OT policy. Confirm the economic buyer who can approve a Rs 2–5 lakh [~] fixed 90-day fee without a full ERP tender.
B2. Prescription card discipline
Every recommendation must name: event, owner, due date, expected rupee line on the DISCOM invoice, operational constraint, and evidence of execution. Cards without owners are deleted. Cap weekly cards at five so plant teams are not flooded. Prefer MD sequencing and idle-hold cuts before exotic tariff products. Never propose changing validated recipe, membrane current, sterile pressure, or Tempcore quench parameters. If quality or safety forbids a schedule change, mark the lever as blocked and move on — credibility beats aggressive savings claims.
B3. Verifying savings without gaming
A lower bill alone is not success if tonnes, pairs, or batches fell. Normalise against a production proxy agreed in week 1. Document holidays, forced outages, new-line commissioning, captive/solar settlement changes, and weather-sensitive HVAC. Keep a decision ledger shared with the plant: recommendation, owner, status, expected vs observed invoice delta. Present kill or expand at day 90 with the same ledger — this is how Stamped differs from audit PDFs that never close.
B4. Competitive and incumbent handling
If the site has EMS/SCADA/BMS, OEM analytics, ISO consultants, or solar EPCs, congratulate the spend and ask what residual bill lines still surprise them monthly. Position Stamped as the closure layer: rupee assignments and invoice reconciliation. Refuse head-to-head dashboard bake-offs. If Gujarat, Panipat, or Noida corporate IT appears, keep plant electrical as the proof owner and let corporate rubber-stamp after a verified bill line — not before.
B5. Geographic and cluster logistics
Batch 3 spans Mandi Gobindgarh, Paonta/Derabassi/Rajpura, Ghaziabad/Muzaffarnagar, and Kanpur–Unnao. Plan field days by corridor. Carry printed one-pagers: Fit score, DISCOM hypothesis, strongest wedge sentence, and two discovery questions. After each visit, update extras field intel — especially corrected phones, electrical names, and whether the bill cleared Rs 30 lakh/month.
B6. Messaging hygiene for this batch
Owner-operated steel/leather accounts tolerate Hindi-friendly WhatsApp and short calls. Listed pharma/chemical accounts prefer email plus LinkedIn with technical depth. Never open with controversies (GST/Excise history, EuGMP observations, rating actions) — those are internal briefings only. Always name DISCOM early in research-backed prep sheets so call-prep UI can render correctly.
6.4 Evidence discipline and next research actions
This dossier separates three evidence classes. Verified public facts are sourced corporate disclosures, official filings, certification material and named releases. Directional operating hypotheses use the disclosed industry/process profile to identify what should be checked; they are not claims about a specific machine, event, bill or control failure. Commercial estimates—especially electricity spend—remain estimates until the site provides a current invoice and production context.
Practical sequence after first contact: (1) verify entity and invoice ownership; (2) collect two PVVNL invoices and document tariff, demand and PF lines; (3) identify the top three load events with the electrical/utility owner; (4) map production constraints that cannot be moved; (5) agree the smallest controllable boundary for a 90-day trial. For any invoice-derived opportunity model, calculate demand, energy, reactive/PF and applicable tariff components separately. Normalise against operating hours or tonnes/batches; a lower bill is not proof of operational savings if production fell. Security and governance should be addressed before data transfer. Treat adverse information proportionately: a negative search does not prove clearance; a regional proceeding is not evidence against an unrelated site. Re-check material sources and contact tenure immediately before a campaign send.