Depth bar: Band B Faridabad visit dossier (~2,000–3,500 words). Process-strong HPDC; do not call ₹30L+ on single meter. Use [~] for estimates. No invented bills or court outcomes.

1. Company overview & snapshot

1.1 Legal identity & corporate structure

Oswal Castings Private Limited (CIN U28998HR2010PTC040998 per public aggregators) — unlisted private company, ROC Delhi/Haryana filings, incorporated 02 Aug 2010. Registered address commonly cited as Plot No. 21, Sector 25, Faridabad 121004. Authorised/paid-up capital reported around ₹10 Cr []. Current directors in aggregator views: Rajan Jain (Managing Director), Raj Kumar Jain (Director), Neeraj Jain (Director). Company secretary historically listed (e.g. Pankaj Kumar Mittal on Tracxn). Open bank charges / financed capex have been material in prior lead notes (₹184 Cr+ charges [] in older CARE-linked summary) — treat as growth leverage signal, not liquidity judgement. Reconfirm MCA directors, charges and GSTIN before proposal.

Entity collision risk: do not confuse with unrelated “Oswal” groups (cables, textiles). Neeraj Jain’s aggregator cross-directorships into textile names need care — verify which Neeraj is on this board before citing group maps.

1.2 What they make & where money comes from

Aluminium high-pressure die cast and precision-machined components for automotive, white goods and general engineering. Website (Aug 2026 check): three plants — Faridabad Unit-1, a second Haryana unit (Palwal modernisation historically), and Pithampur/Indore Unit-III. Unit-1 public equipment picture: central melting furnaces (0.5T / 1T / 3T class), 25 HPDC machines (site copy; lead research used 26 — treat as ~25–26), machine shops, paint shop, full DG backup, ETP, compressors, RO, cooling towers. Revenue: lead report ₹356 Cr FY2025 []; Tofler-style ranges INR 100–500 Cr for FY2024 []. Named OEM exposure in older CARE summary included Tata Motors Passenger Vehicles — do not put OEM names in cold email unless contact confirms.

Commercial logic: OEM/Tier price-downs vs volatile Al and power. Electricity is a margin-defense lever, not an ESG story for the first meeting.

1.3 Plants, addresses & footprint

Confirm exact Palwal/Pithampur addresses, commissioning status and which legal entity holds each HT account. Visit week: Sector 25 only unless Rajan invites a second site.

1.4 Leadership & CRM map

• Rajan Jain — MD / operations — https://in.linkedin.com/in/rajan-jain-5b050b1bb — email rajan@oswalcastings.com (inferred); info@oswalcastings.com verified generic
• Raj Kumar Jain — Director / founder generation
• Neeraj Jain — Director
• Sanjay Joshi — Electrical Maintenance Engineer with energy-project history on LinkedIn — confirm still employed; ideal Path A technical champion after MD warms
• Ask reception: who owns HT bills, sanctioned demand, compressor/furnace electrical maintenance across Faridabad + Palwal

Decision path: Rajan sponsorship → Sanjay/electrical for meters → finance for bills → 60-Day Proof Run on one Faridabad feeder group.

1.5 Recent news (24 months) & timing for Stamped

Strongest signal remains multi-plant expansion (Palwal + Pithampur). That creates both opportunity (need consistent baselines) and risk (commissioning teams prioritise output over analytics). Visit timing: Utso in Faridabad next week with an active local deal at LNM Auto (Sector 59) — useful as a light peer reference (“we’re already working with another Faridabad auto plant”) without disclosing LNM internals. Best window: after a billing cycle when the DHBVN invoice is fresh.

2. Energy profile

DISCOM / supply (name early): Faridabad and Palwal are under DHBVN. Pithampur is a separate Madhya Pradesh supply context — never blend MP and Haryana tariffs into one “group bill” claim.

2.1 Bill band, tariff & demand

Corrected estimate (Aug 2026 — kVA/meter-based): ₹6–15L/unit [~]; ₹15–25L major Faridabad units [~] · Band B · confidence High-medium · ₹30L+ gate: No (single unit) · visit action: Do not call ₹30L+ yet. Canonical: bill-estimates-corrected-aug2026.md.

Working hypothesis for Faridabad-centric / group electricity: ₹6–15 lakh/month [~] per identified unit; ₹15–25 lakh [~] across major Faridabad units — do not call Band A on a single meter (kVA/meter-corrected Aug 2026). Confirm with two recent DHBVN bills per connection. Ask: billed kWh, recorded MD, CMD/sanctioned demand, PF incentive/penalty, ToD if applicable, whether Palwal is separate, fuel vs electric melting.

MD spikes in this plant type often come from furnace operation overlapping HPDC startup and compressor/cooling ramps — not from a single failed motor.

2.2 Generation, fuel & renewables

Full DG backup disclosed on website. That is resilience, not proof of routine DG dispatch, captive generation, solar or open access. Melting may be electric or fuel-heavy depending on furnace design — discovery must split grid levers from fuel levers.

2.3 EnMS, PAT, ISO, BRSR

IATF 16949, ISO 9001, ISO 14001, ISO 45001 publicly claimed. No ISO 50001, PAT DC or BRSR energy table found. Expect pragmatic local energy practice, not a corporate EnMS office.

2.4 Likely ₹ leak categories (hypothesis)

Furnace hold/startup coordination; coincident HPDC + compressor + cooling MD peaks; compressed-air pressure/leaks; cooling-tower/chiller baseload; paint-shop scheduling; idle auxiliaries between shifts; unnecessary DG run if grid is stable. Every item is a hypothesis pending bill + production calendar.

3. Operations, equipment & digital stack

3.1 Process flow & critical loads

Al melt/hold → HPDC cells → trim/machine → paint/finish → inspection/dispatch. Critical electrical/utility loads: furnaces (where electric), HPDC hydraulics/auxiliaries, compressors, cooling towers/pumps, machining centres, paint equipment, DG when used. Map feeders by process block before press-level claims.

3.2 Shifts, seasonality, production pattern

Lead research describes near-24×7 when demand is healthy [~]. Confirm three-shift roster, planned shutdowns, OEM dispatch peaks, weekend hold policy. Shift change and restart windows are the first analysis slices.

3.3 Automation, metering, SCADA/EMS/DCS

Engineering software and advanced HPDC kit suggest manufacturing sophistication; no named EMS/SCADA/historian found. Path A: bills + main/interval meters + any feeder meters + read-only OEM/PLC exports. Path B only if a feeder needs instrumentation — price separately; “no hardware” does not mean “no data work.”

3.4 Capex / tech projects affecting energy

Palwal modernisation and Pithampur capacity change baselines. Compare like-for-like parts and shifts; do not claim “expansion energy problem” as verified fact.

4. Stamped Energy fit analysis

4.1 ICP scorecard

Energy intensity: pass. Geography Faridabad/Palwal: pass (DHBVN). Bill ≥ ₹30L: Fail on single unit [~] — ₹6–15L/unit. Owner-led access: pass. Data maturity: unknown. Multi-plant expansion path: pass. Overall discovery priority: 9/10.

4.2 Fit score rationale

9/10 for process intensity + multi-site follow-on + accessible MD. Reduced from 10 because bill, utility owner and connectivity are unconfirmed. Failed bill gate → nurture / lower commercial motion.

4.3 Wedge (parser-critical)

The strongest wedge is: visit Sector 25 next week, take one Faridabad DHBVN bill boundary and available meters, expose avoidable coincident demand across furnace / HPDC / compressor / cooling startup, assign a production-safe sequencing action, and run a 60-Day Proof Run with results verified with evidence (optional bill confirm) — with LNM Sector 59 as a light local peer if useful.

4.4 Objections & competitors

“We have machine dashboards” → can they link to MD/PF/invoice? “Maintenance handles it” → test one recurring loss. “We have DG/solar” → does not erase demand waste. OEM EMS → Stamped is read-only action + verification layer. Procurement formality at ₹300 Cr+ scale → keep first ask tiny (one plant, 60 days).

4.5 Pilot design

Site: Faridabad Unit-1, one HT connection / bounded feeder group. Offer: 60-Day Proof Run. Success: ranked named-action queue; one ops-approved action; ₹ movement on Stamped ledger vs documented baseline. Kill: bill below gate; no read-only data path; no intervention owner. Then: Palwal compare; Pithampur later with MP DISCOM honesty.

5. Before you reach out

5.1 Discovery checklist

• Confirm Rajan availability for visit week (Sector 25)
• Faridabad-alone vs group bill ≥ ₹30L/mo on DHBVN
• Melting fuel vs electric
• Who owns HT bills centrally vs plant
• Sanjay Joshi still current?
• MD events vs compressor vs furnace pain ranking
• Existing meter/SCADA/OEM dashboards
• Shift roster + weekend hold
• LNM peer mention OK? (yes, light — no confidential ops detail)
• Palwal same legal entity / separate account?
• Any EMS consultant or ISO energy project live?
• Safety veto rules for stagger

5.2 Do not lead with

• Dashboards, AI buzz, ESG-first
• “Next DISCOM bill” as hero proof — use verified with evidence
• Groupwide rollout on day one
• Named OEM customers unless confirmed
• Claiming CARE rating details you cannot cite live

5.3 Opening hooks

• “I’m in Faridabad next week — 20 minutes at Sector 25 on furnace/HPDC MD overlap, 60-Day Proof Run.”
• “We’re already working with another Faridabad auto plant nearby; for Oswal I’d start one DHBVN boundary only.”

6. Risks, flags, controversies & sources

6.1 Integrity / controversy / regulatory (search explicitly)

Aug 2026 refresh: “Oswal Castings controversy / NGT / pollution / lawsuit / fraud / labour” — no substantiated public hits in reviewed sources. Not clearance. Pre-contract: MCA litigation, HSPCB, charge sheet if any.

6.2 Data quality flags

• Private company — limited audited energy disclosure
• HPDC count 25 vs 26 across sources
• Revenue ₹100–500 Cr range vs ₹356 Cr point estimate [~]
• Pithampur vs “Indore” naming inconsistency in older notes
• Sanjay Joshi employment currency unverified

6.3 Sources consulted

• https://oswalcastings.com/ (about, journey, HPDC, unit descriptions)
• https://in.linkedin.com/in/rajan-jain-5b050b1bb
• https://www.linkedin.com/company/oswal-castings-pvt-ltd---india
• Tofler / Tracxn / TheCompanyCheck aggregator pages (CIN U28998HR2010PTC040998)
• leads/auto-components/lead-research-auto-component-NCR-2026-06.md
• Prior: outreach/2026-07-best-prospects-band-a/04-oswal-castings.md + research twin
• Visit context: Faridabad trip + LNM active deal (internal)

6.4 Visit choreography (Sector 25)

Ideal call order for visit week: WhatsApp/email Rajan 48 hours out → confirm 20–30 minutes at Plot 21–23 → ask him to bring electrical lead (or Sanjay Joshi if still there) and two DHBVN bills for Faridabad. On site, do not open with a product deck. Open with the operating question: when last month’s bill surprised them, how long did it take to know whether melting hold, HPDC startup or compressors drove it. Walk the utility spine if permitted — melting, die-cast hall, compressor/cooling, paint — and mark which blocks share the recorded-demand interval. Propose one Faridabad feeder group for the 60-Day Proof Run. Leave Palwal and Pithampur as “later compare,” and say clearly that Pithampur is a different DISCOM so group dashboards must not fake tariff equality.

If Rajan asks who else Stamped works with locally, mention LNM Sector 59 lightly. If he asks for references with HPDC at Oswal’s scale, be honest: LNM is smaller and process-mixed; Oswal’s proof will be its own furnace/HPDC coincidence story. Early-deployment 15–20% is a benchmark from other plants [~], not a promise for Sector 25 until baseline exists.

6.5 Multi-plant data reality

Three plants create three bill cultures. Faridabad may have the richest meter picture; Palwal may still be settling after modernisation; Pithampur may be commissioning-biased toward output. Stamped should refuse a day-one “single pane of glass” commitment. After Faridabad proof, compare intensity metrics (₹/ton or ₹/shot) with production context — not raw kWh. Bank-financed growth (historical charge levels [~]) can make finance sensitive to subscription framing: keep the Proof Run as a fixed, killable verification fee, not an open-ended SaaS argument.

Machine OEM dashboards (Bühler/Toshiba-class if present) often show cycle and alarm data without MD attribution. That objection is common; answer with a request to test one recurring coincident-start week against the DHBVN invoice lines, verified with evidence on Stamped’s ledger.

6.6 Research conclusion

Oswal is the top cold target on this Faridabad list for energy intensity and owner access. Visit week should secure bill copies, a technical champion and a single-plant Proof Run — nothing more. Cross-site expansion is earned after Faridabad evidence, not sold in the lobby.