Reports Funnel diagnostic Pipeline hours Operations diagnostic Fix register

Operations diagnostic — from sanction to first EMI

Real files sanctioned April 2025 to June 2026 (14,461 sanctions, 9,137 followed to payment release), the open PDD/OTC stock on 9 September, the disbursed book on 31 August, the technical, handover, ADR, fee and insurance-booking reports, and the digital-journey MIS. Nine sources, joined on loan number. Rupees at sanction value unless marked paid-out.

14.5 dSanction → payment release, median (p90 56)

9.1 dof it is sanction → docket — 77%

43%of files paid the same day the docket is in

11,484open PDD items on 3,645 loans, median 209 d old

₹99 Crsanctioned, undisbursed on partially-disbursed loans

0 dPayment → handover to seller, median; 13% over 7 d

Seven findings, in order of money

Security is not perfected on 3,645 disbursed loans. 11,484 post-disbursement documents open, median 209 days after the money went out, 60% over 90 days, 20% over a year. 1,336 loans carrying ₹105 Cr outstanding are missing the registered mortgage deed or the EC that proves it. Tamil Nadu closes in 14 days; AP has 94% and Chhattisgarh 86% of items past 90 days. This is the largest risk item in the study and it has no owner above the branch.
Sanction → docket is 77% of the post-sanction clock and none of it is ops. 9.1 days median to get the customer to sign; 5,271 sanctions (₹543 Cr) died with zero activity after the letter — 249 of them ever reached docket. Once the docket is in, ops pays 43% the same day and 72% within five. The chain works; the customer is not committed when the letter is issued.
₹99 Cr is sanctioned, documented and not paid. 2,098 partially disbursed loans; ₹37 Cr with no tranche for more than 90 days; 300 loans with ₹12 Cr past a year. Construction takes 143 days first-to-last tranche. No ageing report, no owner per file.
One branch step accounts for the ops tail. Docket printed → OTC documents collected is 0.9 days median but 21% of elapsed time; Gujarat runs it at 9.9 days (p75 21), CG 11, North 3, South 0.1. The HO query loop on the same step resolves in 1.5 hours. The branch is not collecting until chased, and a printed cheque waits in a drawer.
Ops re-verifies the file three times and the rails that would stop it are switched off. 2,997 ops-raised ADRs at docket and 1,955 HO OTC queries, 85–87% answered within a day, asking for KYC, signatures, NACH, payee accounts and sanction conditions. The LOS has e-Sign, e-NACH, penny-drop and Account Aggregator wired in; on 550 recent files their use is 0.2%, 0.5%, 13% and 0.3%.
Re-sanction is where sales and credit renegotiate at ops' table. 46% of sanctions are re-issued, 77% of them before docket, median 6 days after the first; 23% of re-sanctioned files never disburse. The first letter is not the offer.
Everything after payment is unstamped. Cheque/RTGS, handover to seller, CERSAI, e-NACH, PDD receipt, policy issuance and welcome call have no timestamp in the LOS; two of them exist as snapshot reports, four not at all. Insurance is booked by re-keying each loan into up to three of ten insurer templates a week, with loan-number match rates of 64–100%. The post-payment half of ops cannot be managed by anyone above the branch because it cannot be seen.

What works: docket → release in minutes, checkers under three hours, HO OTC under two, insurance booked within a day of release, 97% insurance attach.

The structural finding

Ops is measured on the part it controls — docket to release, the best-run chain in the company — and blamed for the parts it doesn't: an unsigned customer at docket, documents the branch hasn't collected, a mortgage deed the sub-registrar hasn't returned. The structural finding is that ops has no ownership boundary: it inherits sales' uncommitted sanctions and hands nothing measurable to servicing at first EMI. Every ops "query" system — docket ADR, handover query, PDD list — is one checklist run as a conversation. And the post-payment stages exist as work but not as a process: no timestamps, no ageing, no owner per item. Define ops as the owner of "committed customer → perfected loan, handed to servicing", with an entry gate (commitment at sanction) and an exit gate (PDD complete, NACH live, CERSAI filed), and measure it on that.

The first thirty days

Two numbers move in 30 days on data that already exists: PDD ageing by territory ops manager, weekly, MODT/EC items first, Tamil Nadu's 14 days as the target; and docket-printed → OTC in Gujarat, CG, DNCR and UP under a print-on-OTC rule. Days 30–90: docket checklist in the app, tranche queue on the SUD dashboard, e-Sign and e-NACH made mandatory at docket, and the four missing timestamps into LOS so days 60–90 are measured rather than inferred.

0 The operations pipeline — where sanctioned loans go

Real files sanctioned April 2025 to June 2026 (old enough to have finished), followed through every post-sanction stage the data records. The last two bars are the post-payment stock as it stands today: loans with no open post-disbursement document, and loans with no tranche pending.

Sanctioned

14,461

100% of sanctions

−5,095 never docketed

Docket printed

9,366

65% of sanctions

Docket / Ops done

9,375

65% of sanctions

Disb auth (checker)

9,365

65% of sanctions

−224

OTC docs done

9,141

63% of sanctions

Ops OTC done (HO)

9,140

63% of sanctions

Payment released

9,141

63% of sanctions

−51 no handover date

— after payment: stamped only in snapshot reports, or not at all —

CERSAI created

not stamped — charge collected on every file, filing date unknown

no data

Physical file inward at CPU → checked → stored

not stamped — Section 9A defines the register

no data

Handover to seller / builder recorded

9,090

63% of sanctions

−2,170 with open PDD

PDD complete (no open item today)

6,971

48% of sanctions

1,447 still partially disbursed

Fully disbursed (no tranche pending)

7,694

53% of sanctions

Drop	Loans	Where they sit now	Largest reason in the data	Fix	Owner
Sanction → docket	5,086	Soft Cancelled 83%, Soft Rejected 11%	Customer never signed — zero activity after the letter on almost every one; sanction issued before commitment	Commitment gate before the letter; ops refuses to print a docket without it	Sales + Ops
Docket → payment	236	Soft Cancelled 70%, Cancelled 17%	Dropped at OTC or re-sanction; a few still open	OTC checklist with due date at docket	Branch ops
Payment → handover	51	No handover date recorded	Recording gap rather than a loss	Handover confirmation stamped at branch	Branch ops
Payment → PDD complete	2,170	Open PDD items, median 209 days old	Registered MODT, EC, vetting report, original deed not collected; no owner above the branch	Weekly PDD ageing by TOM, MODT first, due date per item	Zonal ops
Payment → fully disbursed	1,447	Partially disbursed; ₹99 Cr undisbursed on the book	Construction tranches with no site-visit or ageing trigger	Tranche queue on the SUD dashboard with RM named	Ops + Sales

Between sanction and payment the pipeline loses 5,086 loans before docket and 236 after — 96% of the loss is before ops touches the file. Between docket and payment ops loses almost nothing and takes almost no time. After payment the picture reverses: of 9,141 loans paid, 2,170 still carry an open post-disbursement document and 1,447 have money left to pay out, and neither is a stage the LOS can see.

The twenty-three stages, and which ones the data can see

Ordered empirically from timestamps, not LOS labels. Payment authorisation is the last stamped event before money leaves; everything after it is either in a snapshot report or unstamped.

#	Stage	Actor	Median	p75	Stamped?	Note
1	Sanction	Credit	—	—	Yes
2	Docket printed	Branch ops	~1 d	—	Yes — field labelled `chq_printing_done`	Label is wrong since FY25
3	Docket / Ops done (agreement, sanction conditions)	Branch ops (117)	9.1 d from sanction	23 d	Yes	77% of sanction → release time
4	Disb auth (checker)	Regional ops (21)	1 h	3 h	Yes
5	OTC docs done	Branch ops (122)	0.9 d	5 d	Yes	Gujarat 9.9 d; 21% of time
6	Ops OTC done	HO ops (8)	1.5 h	3 h	Yes
7	Payment auth 1 → 2 (release)	Kotkar / Pradhan / System → Deousker / Kotkar / Dhiman	minutes	—	Yes	Sales and Insurance sign the money out
8	Cheque / RTGS	HO / Treasury	same day 72%	1 d	No — LDR First Disb Date only
9	Handover to seller / builder	Branch	0 d	2 d	LDR only	13% > 7 d; construction 13 d
10	CERSAI creation	Ops	—	—	No	Charge collected at initiation; filing date, asset ID and status unstamped
11	E-NACH registration → active	Ops / customer	—	—	No	386 of 388 sanctioned files "Not Initiated"
12	Insurance booking → policy issued	Ops → 10 insurers	1 d to fund transfer	—	Booking files only	No policy number or issue date
13	Welcome call / kit	Ops / CS	—	—	No
14	Physical file made at branch — outward to CPU	Branch ops	—	—	No	Dispatch date, courier / bag reference, sender
15	Physical file inward at CPU	Central processing	—	—	No	Receipt date, receiver; transit days
16	CPU check — checker, date, discrepancy	CPU checker	—	—	No	Checklist result per document; discrepancy code
17	CPU query raised → resolved	CPU ↔ branch	—	—	No	Query text, raised / resolved date, resolver
18	File to storage — vault / location	CPU / records	—	—	No	Storage location, box reference, date
19	PDD — originals, MODT, EC, vetting	Branch ops / Legal	209 d median age	359 d	PDD report only (snapshot)	11,484 open items
20	First EMI	Customer	43 d after disbursal	51 d	LDR	Hand-off to servicing; outside this study
21	Subsequent tranches	Branch ops → HO (repeat)	143 d first → last	222 d	Yes, as latest_ stamps	₹99 Cr undisbursed
22	Foreclosure / closure → NOC	Ops	—	—	LDR foreclosed date only	NOC issue date unstamped
23	Document return — outward from storage → branch → customer	Records / branch	—	—	No	Retrieval date, dispatch, branch inward, handover, customer acknowledgement

Measured from the LOS event log, August 2025 to September 2026

Milestone gap — event log, real files, Aug 25–Sep 26	Median d	p75	p90	Files
Login → QC	0.2	1.2	4.0	30,163
QC → U1	3.0	6.2	14.0	18,913
U1 → Biz UW	0.9	3.0	8.2	17,384
Biz UW → U2	4.1	9.9	20.3	13,528
U2 → sanction	0.0	0.6	1.7	13,206
Login → sanction	14.0	23.8	39.9	13,745
Sanction → docket printed	7.9	19.1	38.1	8,716
Disb auth → OTC done	0.7	3.8	12.0	8,412
OTC → OTC compliance done	0.1	0.1	0.7	8,403
OTC compliance → payment	0.0	0.0	0.1	8,405
Sanction → disbursed	12.0	27.0	48.0	8,376
Login → disbursed	31.4	51.9	77.3	8,374

Every completed transition in the LOS from August 2025 to September 2026, recomputed from timestamps. The dump-derived figures above hold to within a day; the event log adds the QC → U1 wait and confirms that U2 → sanction and the payment chain are hours, not days.

Two labels are wrong in the dump: chq_printing_done is docket printing since FY25 (it was cheque printing until FY24 and fell to the wrong side of payment authorisation the year the process changed); the QC report's "Tech Initiation Date" is the LOS stage mark, set after the vendor has already been working for a week. Eleven of the twenty-three stages have no timestamp anywhere — CERSAI, e-NACH, welcome call, the entire physical-file custody chain (dispatch, inward, check, query, storage) and the document-return chain on closure. Two exist only as snapshots. Section 12 has the field spec.

1 Sanction → docket: the 77%

Sanctions Apr 2025–Jun 2026, mature. "Lost" = Cancelled, Soft Cancelled, Rejected or Soft Rejected today. The front-end report sizes this leak; here it is by region, branch and RM because docket is where ops inherits it.

Zone	Region	Sanctions Apr 25–Jun 26	Lost after sanction	Sanction → docket median d	p75	Re-sanctioned	Action
SOUTH	AP	2,043	63%	20	39	16%	No sanction letter without fee / e-sign / Accept-offer; weekly named-file call on every sanction > 15 days
NORTH	DNCR	1,292	42%	9	22	52%	No sanction letter without fee / e-sign / Accept-offer; weekly named-file call on every sanction > 15 days
NORTH	RAJASTHAN	1,343	39%	7	20	50%	Weekly SUD call on sanctions > 15 days; RM named per file
WEST 2	GUJARAT	1,286	37%	9	19	55%	Weekly SUD call on sanctions > 15 days; RM named per file
NORTH	UP	776	36%	6	14	48%	Weekly SUD call on sanctions > 15 days; RM named per file
WEST 1	MMR	1,454	34%	10	27	62%	Weekly SUD call on sanctions > 15 days; RM named per file
WEST 1	VIDARBHA	1,037	34%	14	30	52%	Weekly SUD call on sanctions > 15 days; RM named per file
WEST 1	WEST MAHA	1,192	34%	17	33	53%	Weekly SUD call on sanctions > 15 days; RM named per file
WEST 1	CG	245	33%	16	31	60%	Weekly SUD call on sanctions > 15 days; RM named per file
WEST 1	MP	889	23%	9	21	40%	At reference — hold
SOUTH	TG	563	23%	5	10	38%	At reference — hold
SOUTH	KARNATAKA	1,552	22%	7	20	55%	At reference — hold
SOUTH	TAMIL NADU	786	19%	4	9	40%	At reference — hold

The lost files are not slow — they are dead on arrival. Median days of any activity after sanction on a lost file: zero. 249 of 5,271 reached docket; 26 reached payment. The fix is a commitment gate before the letter (fee, e-Sign or Accept-offer reply), and it belongs to sales; ops' part is to refuse to print a docket for a sanction with no commitment on file, which is what a gate is.

Branches and RMs — worst, then the reference

Region	Branch (≥60 sanctions)	Sanctions	Lost after sanction	Sanction → docket median d
AP	KURNOOL	150	79%	11.0
AP	TIRUPATI	263	75%	6.5
AP	RAJAHMUNDRY	600	67%	26.5
AP	VIJAYAWADA	484	60%	22.0
AP	VIZAG	355	57%	19.0
RAJASTHAN	JODHPUR	115	56%	17.5
RAJASTHAN	AJMER	66	52%	9.0
DNCR	MEERUT	99	48%	9.0
UP	VARANASI	93	47%	6.5
DNCR	NOIDA	138	47%	6.0
DNCR	FARIDABAD	142	46%	7.0
GUJARAT	JUNAGADH	76	46%	9.0

Region	RM (≥25 sanctions in window)	Sanctions	Lost after sanction	Never reached docket	Sanction → docket median d
AP	V Suresh Babu	37	92%	92%	35.0
AP	Venu E	32	84%	78%	15.0
AP	Santosh Thavalam	39	80%	80%	4.0
AP	Narendar Boya	34	79%	79%	11.0
AP	Venkatesh satyavada	50	76%	74%	43.0
AP	Surendra Madasu	34	74%	74%	23.0
AP	Munivasantharao G	37	73%	73%	9.5
AP	Anwar Basha Shaik	53	70%	68%	20.0
AP	Saravana Kumar P	26	69%	69%	8.0
AP	Gopinadh Velpula	69	68%	71%	18.0
AP	Sivasankar Pujari	36	67%	67%	5.5
AP	Siva Vennelakoti	57	67%	63%	22.0
AP	Mohan Koppisetti	48	65%	65%	12.0
AP	Venkata Sai	56	64%	62%	25.0
AP	Suryanarayana Namana	64	61%	61%	15.0

200 RMs with 25 or more sanctions in the window; the median RM loses 31% after sanction. The table above is the tail; below is what the same job looks like when it is done.

Region	RM — the reference (≥40 sanctions)	Sanctions	Lost after sanction	Never reached docket	Sanction → docket median d
TAMIL NADU	Igbal S	52	4%	4%	1.0
MP	Vikas Soni	47	6%	6%	11.0
KARNATAKA	Raveesh SJ	43	9%	9%	5.0
UP	Rinku Sharma	106	9%	9%	5.0
WEST MAHA	Sunil Chavan	57	10%	10%	2.0
MMR	Sunil Chavan	57	10%	10%	5.0
MP	Ravindra	61	13%	13%	10.0
WEST MAHA	Akshay Junghare	52	15%	15%	19.0

Re-sanction

6,717 of 14,461 sanctions (46%) were re-issued; 77% before docket, median 6 days after the first letter; 23% of re-sanctioned files never disburse. Terms are being renegotiated at the agreement table after credit has signed. The LOS comment field names the reasons: the largest is a mismatch between the sanction and the technical report (property category, land type) — 7,341 re-sanction transitions on 5,842 loans in the year — followed by ROI and amount changes. Technical initiated at login removes the first; right-first-time sanction — amount, tenor and ROI confirmed with the customer before the letter — turns the rest into an exception path.

2 Docket → payment: the chain that works, and the one step that doesn't

Step	Median days	p75	p90	Share of sanction → release time
Sanction → Docket printed	9.76	23.89	46.02	78%
Docket printed → Docket / Ops done	-0.06	-0.03	-0.01	0%
Docket / Ops done → Disb auth (checker)	0.04	0.12	0.76	1%
Disb auth (checker) → OTC docs done	0.90	5.00	14.17	21%
OTC docs done → Ops OTC done (HO)	0.06	0.12	0.73	1%
Ops OTC done (HO) → Payment auth 1	0.00	0.01	0.03	0%
Payment auth 1 → Payment auth 2 — release	0.01	0.01	0.03	0%

Once the docket is in: 43% released the same day, 57% within two, 72% within five, 16% over ten. In the slow tail every HO step is still hours; the OTC step is ten days. It is one step, and it is the branch's.

Docket printed → OTC done, by region

Zone	Region	Docket printed → OTC done, median d	p75	Docket → release > 5 days	Action
WEST 1	CG	10.2	17.9	46%	Print-on-OTC: cheque only after OTC checklist complete; OTC due date set at docket
WEST 2	GUJARAT	9.9	20.9	44%	Print-on-OTC: cheque only after OTC checklist complete; OTC due date set at docket
NORTH	DNCR	4.9	11.9	28%	Print-on-OTC: cheque only after OTC checklist complete; OTC due date set at docket
WEST 1	VIDARBHA	3.1	8.4	24%	Print-on-OTC: cheque only after OTC checklist complete; OTC due date set at docket
NORTH	UP	2.9	7.4	25%	Print-on-OTC: cheque only after OTC checklist complete; OTC due date set at docket
NORTH	RAJASTHAN	1.9	5.0	14%	Watch
WEST 1	MP	1.0	6.0	22%	Watch
SOUTH	TAMIL NADU	0.1	0.8	3%	Reference — OTC before print already
SOUTH	KARNATAKA	0.1	0.9	4%	Reference — OTC before print already
SOUTH	TG	0.1	0.9	2%	Reference — OTC before print already
SOUTH	AP	0.1	1.0	3%	Reference — OTC before print already
WEST 1	MMR	0.1	1.1	7%	Reference — OTC before print already
WEST 1	WEST MAHA	0.0	0.3	7%	Reference — OTC before print already

South collects OTC before the docket is printed; Gujarat and CG print and wait. The HO handover queries (below) show the branch answers HO in 1.5 hours when asked — so the gap is not the query loop, it is that nobody asks until the cheque is sitting there. Print-on-OTC reverses the order: the OTC checklist gets a due date at docket, the cheque prints when it is complete. 6,946 PDD items today are on loans where the cheque is still marked "in hand".

By person

Disbursement checker (regional ops)	Files	Docket → auth, median	p75
Manjunath M SOUTH	1,620	1.2 h	2.6 h
Ashutosh Sharma NORTH	1,282	1.7 h	16.6 h
Venkata Sivaprasad SOUTH	1,218	1.2 h	2.6 h
Sandeep Gupta WEST 1	852	0.7 h	4.8 h
Pranesh Minde WEST 1	800	0.2 h	1.7 h
Akshay Jadhav WEST 1	786	0.7 h	1.7 h
Pravin Randive WEST 1	692	0.0 h	0.2 h
Mukesh Vasava WEST 2	688	1.0 h	4.6 h
Rahul Sharma NORTH	464	1.2 h	2.9 h
Nayanraj Panwar NORTH	227	2.4 h	10.1 h
Madu C SOUTH	210	1.0 h	2.6 h
Hardik Baraiya WEST 2	167	3.1 h	17.5 h

Branch ops — OTC done by	Files	Docket printed → OTC, median	p75
Abhay Madnani WEST 1	272	0.0 d	2.3 d
Amit Sable WEST 1	257	0.9 d	4.0 d
Vishal Shinde WEST 1	234	0.0 d	1.0 d
Manglesh Sharma WEST 1	233	2.1 d	7.2 d
Dorasanaiah Mekala SOUTH	228	0.1 d	1.1 d
Yakshit Patel WEST 2	202	6.0 d	15.0 d
Pratik Vare WEST 1	201	0.0 d	0.1 d
Baskaran C SOUTH	170	0.1 d	0.7 d
Popuri Gowtham SOUTH	167	0.1 d	0.9 d
Akshay D SOUTH	167	0.7 d	1.7 d
Yash Ramdev NORTH	165	2.1 d	6.9 d
Sumit Agarwal NORTH	161	2.3 d	6.7 d
Amol Kapre WEST 1	156	5.9 d	12.0 d
Ravindar Alukaraju SOUTH	153	0.7 d	3.8 d

The HO OTC query loop (handover report)

1,955 queries on 1,638 loans since Apr 2025, raised by the five HO OTC people to the branch; median 1.5 h, p90 24.4 h, 87% within a day; 65 unresolved, oldest 2025-05-19. Half are raised between 10:00 and 13:00.

HO OTC query to branch (handover report, Apr 25 →)	Queries	Median hours to resolve
Sanction conditions / approvals	538	1.7 h
PDD / originals / MODT / vetting	359	1.4 h
Sanction letter / agreement signatures	297	1.8 h
Fee / charges / stamp	242	1.8 h
KYC / income of applicant	202	3.0 h
Seller KYC / CIBIL / OSV	182	1.3 h
Tranche / construction photos	125	2.9 h
NACH / SPDC / mandate	124	1.5 h
Cheque favouring / RTGS	94	2.8 h
Insurance	65	4.1 h

3 Payment → handover

Product	Loans (disbursed Apr 25 →)	First disbursal → handover of largest amount, median d	p75	> 7 days
Home Purchase	3,945	0	0	2%
Home Construction / Extension	2,063	13	58	54%
LAP	1,810	0	0	6%
Top Up	1,203	0	0	7%
Balance Transfer	684	0	0	1%
Plot Purchase + Construction	525	0	0	6%
Micro LAP	335	0	0	4%

Non-construction handover is same-day; the 13% over seven days on Home Purchase is the customer paying interest on money the seller has not received. Construction is tranche-driven by design. First EMI falls 43 days after disbursal (p75 51); what happens from there is servicing and is out of scope here.

4 PDD — security perfection

Open post-disbursement items on 9 September. Age is days since first disbursal. Outstanding is the loan balance on the loans carrying open items.

Age distribution of the 11,484 items:

≤30 d	31–90	91–180	181–365	> 365
1,617	1,110	1,292	3,290	2,341

Region	Loans with open PDD	Items	Outstanding ₹Cr on those loans	Median age (d since disbursal)	> 90 d	> 1 yr	MODT / EC items	Action
ANDHRA PRADESH	424	1,757	16	219	94%	0%	913	Escalate: TOM to clear MODT/EC items first; branch cannot originate new PDD-heavy cases until under 50%
CHHATTISGARH	269	581	9	561	86%	75%	12	Escalate: TOM to clear MODT/EC items first; branch cannot originate new PDD-heavy cases until under 50%
HARYANA	43	224	5	348	85%	44%	2	Escalate: TOM to clear MODT/EC items first; branch cannot originate new PDD-heavy cases until under 50%
DELHI	133	823	15	338	65%	31%	3	Weekly PDD call, MODT items named; 60-day target to Tamil Nadu standard
KARNATAKA	401	1,530	42	173	63%	13%	500	Weekly PDD call, MODT items named; 60-day target to Tamil Nadu standard
MAHARASHTRA	1,323	2,873	173	197	58%	27%	308	Weekly PDD call, MODT items named; 60-day target to Tamil Nadu standard
MADHYA PRADESH	336	803	25	167	53%	18%	43	Weekly PDD call, MODT items named; 60-day target to Tamil Nadu standard
UTTAR PRADESH	288	991	32	220	52%	22%	12	Weekly PDD call, MODT items named; 60-day target to Tamil Nadu standard
GUJARAT	223	1,174	12	187	33%	13%	218	Watch; MODT items to zero in 30 days
TELANGANA	55	148	8	74	26%	3%	69	Watch; MODT items to zero in 30 days
RAJASTHAN	78	269	7	18	17%	16%	44	Hold at standard
TAMIL NADU	69	307	12	14	5%	0%	105	Hold at standard

By document

Document type	Items	Loans	Median age d	> 90 d
Other	2,884	1,580	175	44%
Registered MODT	2,140	1,317	204	72%
Original sale deed	1,988	1,660	222	64%
Vetting report	1,810	1,793	227	75%
Possession / khata / NOC	1,066	596	263	76%
OCR receipt	712	495	191	42%
KYC / address	433	339	199	58%
SPDC / NACH	362	288	57	28%
EC reflecting MODT	89	71	15	19%

Registered MODT and the EC reflecting it are the mortgage. 1,336 loans, ₹105 Cr outstanding, median 201 days without either, 68% over 90 days. Vetting reports are the lawyer's; the "OTC / PDD of" field says 14% of all items are the lawyer's, 6% EHFL's own. This is an audit and NHB inspection exposure before it is a TAT.

By territory ops manager

Territory ops manager (≥150 items)	Region	Loans	Items	Median age d	> 90 d
Aakash Tomar	UTTAR PRADESH	76	317	313	100%
Amit Sable	MAHARASHTRA	134	153	1291	100%
Ravindar Alukaraju	ANDHRA PRADESH	105	428	222	100%
Somnath Nishad	CHHATTISGARH	178	349	691	100%
Pooja S	KARNATAKA	53	245	345	100%
Dorasanaiah Mekala	ANDHRA PRADESH	117	490	217	97%
Ashok Thota	ANDHRA PRADESH	46	159	230	96%
Appalaraju Dalai	ANDHRA PRADESH	125	557	202	87%
Pradeepa Kumar	KARNATAKA	52	228	326	87%
Dalip Kumar	HARYANA	42	223	348	85%
Ajay Sodaye	MAHARASHTRA	128	246	208	75%
Sajan Singh	DELHI	56	375	336	73%
Madhugowda T	KARNATAKA	50	165	176	64%
Anshul	DELHI	28	241	299	64%
Santosh Mahakud	MAHARASHTRA	86	190	169	62%
Prashant Umale	MAHARASHTRA	220	494	194	62%
Abhay Madnani	MADHYA PRADESH	142	305	146	61%
Amol Kapre	MAHARASHTRA	94	279	215	59%
Pradeepa M S	KARNATAKA	78	408	114	59%
Ganesh Nikam	GUJARAT	46	188	332	53%
Bharat Bajetha	DELHI	52	211	336	52%
Yash Ramdev	UTTAR PRADESH	88	212	168	46%
Manglesh Sharma	MADHYA PRADESH	107	281	126	44%
Yakshit Patel	GUJARAT	80	377	132	38%
Umesh Wankhede	MAHARASHTRA	33	163	41	26%
Pratik Vare	MAHARASHTRA	84	224	40	19%

Three TOMs have every item past 90 days. The report exists and is exported weekly; what does not exist is a number per TOM on a dashboard, a due date per item and a consequence. Tamil Nadu's 14-day median is the same job under the same LOS.

5 Tranches — the stage nobody tracks

Zone	Partially disbursed loans	Undisbursed ₹Cr	≤30 d since last tranche	31–60 d since last tranche	61–90 d since last tranche	91–180 d since last tranche	181–365 d since last tranche	>365 d since last tranche	Idle > 90 d ₹Cr
NORTH	369	14.9	3.9 104	1.7 39	0.7 21	1.8 53	2.4 53	3.5 77	7.8
SOUTH	682	34.1	10.2 199	6.5 131	4.2 62	5.6 121	3.7 76	1.3 32	10.5
WEST 1	943	46.3	14.2 242	7.2 120	3.4 79	7.0 148	4.5 114	6.4 174	17.8
WEST 2	102	3.8	1.1 29	0.5 11	0.7 10	0.4 19	0.5 13	0.3 15	1.2

Product	Loans	Undisbursed ₹Cr	Median days since last tranche	Median completion %
Home Construction / Extension	1,150	50.4	41	20.0
Plot Purchase + Construction	466	27.8	123	0.0
Home Purchase	259	14.8	102	70.0
Balance Transfer	79	2.6	40	71.7
LAP	76	2.4	70	100.0
Top Up	53	1.0	125	100.0

Construction files take 143 days first-to-last tranche (p75 222); a Plot + Construction file idle for 123 days is normal if someone has visited the site. Nothing here records a visit. The queue belongs on the SUD dashboard as its own tab: stage of construction, days since last tranche, last site visit, RM named, P1–P5 by ₹ and age.

6 Technical — the vendor scorecard

20,786 window logins with a vendor initiation date. Login → vendor initiated 3.2 d median (p90 18.3); initiation → report 3.0 d (p90 17.0); visit within 1.0 d; 32% initiated before U1.

Technical vendor (≥200 files)	Files	Initiation → report, median d	p75	Initiation → visit d	Action
A.B ASSOCIATES NORTH	477	1	2	1	Reference
ANANDAM ASSOCIATES NORTH	527	1	2	1	Reference
GREENFINCH REAL ESTATE ENGINEERS & CONSULTANTS P SOUTH	205	1	2	1	Reference
MOHAN KUMAR D (MOHAN KUMAR AND ASSOCIATES) TECHN SOUTH	420	1	2	1	Reference
M/S MAHALXMI CONSTRUCTION NORTH	298	1	2	1	Reference
MRUNAL LOTHE WEST 2	219	2	6	1	Reference
MOHNISH KUMAR WEST 1	293	2	6	1	Reference
GREENFINCH REAL ESTATE ENGINEERS & CONSULTANTS P NORTH	296	2	4	1	Reference
RAHUL DAINA WEST 1	797	2	6	2	Reference
VALU WIZARDS (HONNAIKANAHAL RAMARAJU ARJUN) SOUTH	252	2	3	1	Reference
PRIYANSHU NORTH	402	2	5	1	Reference
JANASARAN ASSOCIATES SOUTH	220	3	5	1	Hold; watch p75
GREENFINCH REAL ESTATE ENGINEERS & CONSULTANTS P WEST 1	243	3	10	1	Hold; watch p75
GOPINATH SOUTH	259	3	6	1	Hold; watch p75
EEPAC NORTH	641	3	6	1	Hold; watch p75
JAYESH CHAULKAR WEST 1	436	3	11	2	Hold; watch p75
VED SHARMA NORTH	244	3	5	2	Hold; watch p75
RAZIO ENGINEERING CONSULTANCY SEVICES SOUTH	259	3	4	1	Hold; watch p75
SAGAR SHARMA NORTH	267	3	6	1	Hold; watch p75
KANASSURE CONSULTANCY SERVICES PRIVATE LTD. NORTH	718	3	5	2	Hold; watch p75
DHYEY ASSOCIATES WEST 2	211	4	8	1	Hold; watch p75
AMOL MAGAR WEST 1	416	4	12	3	Hold; watch p75
AMAN SHAHU WEST 1	543	4	8	1	Hold; watch p75
HAMSAVISWANATH CHINTHAGINJALA SOUTH	256	4	14	2	Hold; watch p75
EVALU8 APPRAISAL WEST 2	216	4	6	2	Hold; watch p75
AKSHAY DESHPANDE WEST 1	442	6	13	3	Re-tender or cap allocation; 3-day SLA
VEDANT ASSOCIATES WEST 1	294	6	13	3	Re-tender or cap allocation; 3-day SLA
JINAY NATE WEST 1	414	6	25	4	Re-tender or cap allocation; 3-day SLA
SUDHARSHAN NAYBAL WEST 1	539	6	18	2	Re-tender or cap allocation; 3-day SLA
System SOUTH	1,473	13	14	14	Re-tender or cap allocation; 3-day SLA
INTERNAL SOUTH	425	14	14	24	Re-tender or cap allocation; 3-day SLA

Zone	Login → technical initiated, median d	Initiation → report, median d
NORTH	2.4	2.0
SOUTH	3.4	2.0
WEST 1	3.3	4.0
WEST 2	3.4	4.0

Internal valuations ("System" / INTERNAL, ~1,900 files) take 13–14 days; the best external vendors take a day. Boundary mismatch is recorded as "NO" on 3,771 of 20,443 reports — 18%. Dual valuation on 1.4% of files. 9% of collateral has no plan approval. Lat/long is present on 99% of properties — the Collateral Radar input already exists.

7 Ops queries — one checklist, three systems

Ops-raised ADRs between sanction and payment: 2,997 on 2,618 loans, 1.14 rounds per file, 67% same-day, 86% within a day.

What ops asks for between sanction and payment (7,378 ADRs)	ADRs	Answered ≤ 1 day	Replace the question with
Sanction conditions / approvals	1,911	81%	Conditions list auto-carried into docket checklist; approval status visible before print
KYC / income re-asked	1,905	79%	Already in the file — verified at login; ops confirms, does not ask
Technical / legal clearance	1,666	64%	Report status gate: docket cannot print without approved technical + legal
Co-applicant / guarantor	1,530	82%	Co-applicant e-KYC at login (front-end Rule 4)
Original docs / MODT / PDD	1,027	68%	PDD list generated at docket with due dates; not a query
NACH / mandate / SPDC	863	82%	e-NACH at docket — rail exists, 0.5% used
Agreement / signatures / KFS	851	76%	e-Sign — the rail exists and is unused
Insurance	779	84%	Premium and nominee captured at docket; waiver rule, not approval
Cheque favouring / RTGS	664	78%	Penny-drop on payee account at docket; rail exists, 13% used
Fee / stamp / franking	615	73%	Fee ledger status auto-checked before print

The event log counts 3.9 ADR entries per logged loan across the whole journey and 7.1 mean ADR-days; the docket queries below are the ops share of that. Ops runs the tightest query loop in the company and it is still the wrong tool. 85% answered within a day means the branch had it; 1,905 KYC re-asks and 1,530 co-applicant asks at docket are the same documents the login desk asked for in the front-end report. The digital rails that end most of these are already in the LOS (Section 8).

8 The rails exist and are switched off

Digital check in LOS (MIS, 550 QC-approved loans since April)	Completion	What it would remove
e-KYC	93%	Works
PAN verification	0%	PAN re-asks at docket and OTC
Crime check	13%	RCU queries
Penny-drop payee verification	13%	Cheque-favouring / RTGS queries (758 across both loops)
Account Aggregator	0.3%	Bank-statement re-asks; 18,202 login ADRs upstream
e-Sign	0.2%	Agreement / sanction-letter / KFS signature queries (1,148 across both loops)
e-NACH	0.5%	NACH / SPDC queries (987); manual mandate registration

This is a subset (loans that touched the digital journey) and should be re-run on all logins since April. Even so: the tech spec in the front-end report assumed these had to be built. They have to be made mandatory at the stage where the query is currently raised — e-Sign and e-NACH at docket, penny-drop at docket for the payee, AA at login.

9 Insurance booking

Week of 8–11 September	Value	Read
Insurers booked	10	3 life, 4 health, 3 property — each in its own upload template, 27 to 79 columns
Booking rows / loans	281 / 76	Up to three policies per loan, re-keyed from LOS by hand into each template
Loans released in the week / appearing in a booking file	127 / 70	Book attach is 97%; the remainder is either a later batch or booked outside these files — reconcile weekly
Loan-number match to LOS	Bajaj 28/28 · IndusInd 44/44 · HDFC Life 50/66 · Religare 55/75 · TATA AIA 25/39 · Magma 14/21	A wrong loan number in the insurer's file is a claim dispute later
Premium as % of disbursed (book) / median per loan	5.2% / ₹48,552

One canonical booking record generated at payment release, ten templates produced from it, loan number and dates validated against LOS before the file leaves. Policy number and issue date must come back and be stored — today "premium collected → policy in hand" cannot be measured.

9A Physical file custody and document return — the process and the register it needs

None of the nine sources records the physical file. The original agreement, the OTC and PDD originals and the title documents move from the branch to central processing and into storage, and back out on closure — and the LOS does not know when, through whom, or in what condition. This section defines the stages and the register so the next version of this report can measure them.

#	Step	Actor	Timestamp to capture	Also capture	Measure
1	Physical file assembled at branch	Branch ops	File-complete date	Document checklist ticked per item; missing items flagged as PDD	Payment release → file complete (days)
2	Outward from branch	Branch ops	Dispatch date and time	Courier / bag / POD reference; sender; number of files in bag	File complete → dispatch
3	Inward at central processing	CPU inward desk	Receipt date and time	Receiver; bag reference matched; files short or excess	Transit days by branch and courier
4	Check at CPU	CPU checker	Check start / complete	Checker name; checklist result per document; discrepancy code (missing, unsigned, mismatch, expired, illegible)	Inward → checked; discrepancies per file; per checker
5	Query raised to branch	CPU checker	Raised date	Query text; discrepancy code; document	Queries per file; by branch, by discrepancy code
6	Query resolved	Branch ops / RM	Resolved date	Resolver; resolution (document sent, waiver approved, corrected)	Hours to resolve; open queries by age
7	File to storage	Records	Stored date	Storage vendor / vault; box and shelf reference; retrieval reference	Checked → stored; files not in storage after 30 days
8	Foreclosure / closure	Ops	Closure date; NOC issued	NOC reference; closure type	Closure → NOC (days)
9	Retrieval from storage	Records	Request date; retrieved date	Requester; box reference	NOC → retrieval
10	Outward to branch	Records	Dispatch date	Courier / POD	Transit
11	Inward at branch	Branch ops	Receipt date	Receiver; document list verified against storage list	Documents short on return
12	Handover to customer	Branch ops	Handover date	Customer acknowledgement (signed list or OTP); ID verified	Closure → documents in customer's hand — the regulatory clock

Two registers, one loan number. A Docket Movement Register (steps 1–7) and a Document Return Register (steps 8–12), both keyed on loan number, both with actor and timestamp on every row and a discrepancy code on every check. Built as LOS stages, they give the same visibility the docket → payment chain has today; built as spreadsheets, they give nothing. The regulatory point on return is not optional: title documents due back to the customer within the prescribed period after closure, with a penalty for delay — closure → handover is the number an inspector will ask for, and today it cannot be produced.

What the existing data hints at. The handover query log is already the CPU-to-branch query loop for the OTC stage — 1,955 queries, 1.5-hour median resolution, the five HO OTC people raising them. Extending that mechanism to the physical file (steps 5–6) is a configuration change, not a build. The PDD/OTC report's "cheque in hand" flag and the Loan Detail Report's foreclosed date are the only current anchors for steps 1 and 8.

10 Controls

Open post-sanction stock today	Files	₹Cr	Median idle d	Idle > 15 d
sanctioned, no docket	1,237	170.5	11	528
docket printed → OTC	141	23.0	9	51
awaiting payment auth	29	5.3	11	10
docket → disb auth	9	1.1	1	0
OTC done → HO	2	0.3	6	0
disb auth → OTC	2	0.2	1128	2

Control point	Data	Action
Payment authorisation signatories	First authoriser: Sachin Kotkar (Insurance) 3,452, Sunil Kumar Pradhan 2,447, System Admin 1,819, Vinayak Deousker (CBO) 1,312. Second: Vinayak Deousker 3,374, Sachin Kotkar 2,070, Ashish Dhiman 1,622	Money-out signatures move to Finance and Ops; Sales and Insurance off the payment chain. "System Admin" as a signatory is a finding on its own
OTC compliance authoriser	22,780 first-authoriser events in the year on 8,451 loans — 9,376 by one person and 4,073 by a second (event log)	A single-person gate on every disbursal; name a deputy and set a same-day SLA
Cheques in hand	6,946 open PDD/OTC items sit on loans where the disbursement cheque is still marked "in hand" — 1,463 of them OTC (pre-handover)	Print-on-OTC; cheque-in-hand ageing on the daily read
Stale statuses	151 files in "ADR after sanction", median idle 389 days; 65 handover queries open since 2025-05-19	Close; they inflate sanctioned stock
Processing-fee waivers	0% of PF rows waived in the window; Gujarat 0%, Maharashtra ?%, Tamil Nadu 0%	Waiver authority and reason code by state on the monthly review
Label drift in LOS	`chq_printing_done` = docket printing since FY25; "Tech Initiation Date" set after the vendor visit	Rename; every report built on these is wrong by a stage

11 People and hours

Ops stage	09	10	11	12	13	14	15	16	17	18	19	20	21	after 18h	after 20h	Saturday
Docket / Ops done n=8,699	2	7	8	9	8	9	11	12	13	12	5	2	1	20%	3%	16%
Disb auth n=8,691	2	5	10	9	8	9	11	11	13	12	4	2	1	20%	4%	16%
OTC done n=8,520	2	5	7	8	8	10	14	13	14	12	5	2	1	20%	3%	17%
Ops OTC (HO) n=8,521	0	3	7	7	4	9	14	13	13	17	8	3	1	29%	5%	17%
Payment auth 2 n=8,519	1	1	5	8	5	6	12	14	13	18	9	4	2	33%	6%	17%

Payment authorisation and release peak in the 18h hour and 17–19% of ops volume is Saturday. A 13:00 batch moves a third of releases a day earlier; branch OTC and docket work is spread through the afternoon and follows credit's evening output. Ops has already moved out of the night (FY22: 51% of authorisations after 20h) — the remaining change is the second batch.

12 Data — what to add and what to fix

Rename: chq_printing_done → docket_printing_done; QC "Tech Initiation Date" → "Tech stage marked"; add the vendor initiation date from the Property report to the dump.
Add timestamps (with actor): cheque / RTGS UTR time; handover-to-seller confirmation; CERSAI filing date, asset ID and status; e-NACH mandate registration and activation; PDD item created / due / received / waived; policy number and issue date from each insurer; welcome-call done.
Physical file custody (Section 9A): file-complete, branch outward, CPU inward, check start/complete with checker and discrepancy code, query raised/resolved with resolver, storage location and date — the Docket Movement Register.
Document return (Section 9A): closure, NOC issued, retrieval requested/retrieved, outward, branch inward, customer handover with acknowledgement — the Document Return Register.
Add fields to the dump query: Employment Type, Income Program, APF flag, tranche number and amount, PAN hash for dedupe.
Export in full: the Loan Workflow report (every status transition with actor and TAT) for the whole book — it replaces most of the inference in this report with measured queue time per status per person. The 10-day sample already lists 18 transient states between sanction and disbursal.
Re-run on all logins: the digital-journey MIS (e-KYC, PAN, AA, penny-drop, e-Sign, e-NACH) so Section 8 is the book, not a sample.

13 Ninety-day plan

#	Move	Owner	Metric and target	Worth
1	PDD ageing by TOM on a weekly dashboard; MODT/EC items first; due date per item; AP and CG escalated to zonal ops	COO + zonal ops heads	Items > 90 d: 60% → 25% by Dec; MODT/EC open loans 1,336 → under 300	Security on ₹105 Cr
2	Ops refuses a docket for a sanction with no commitment on file (fee / e-Sign / Accept-offer); sanction → docket by RM on the SUD dashboard	CBO + COO	Post-sanction loss 36% → 25%; sanction → docket 9 → 5 d	~₹12 Cr / month
3	Print-on-OTC: cheque prints when the OTC checklist is complete; OTC due date set at docket; Gujarat, CG, DNCR, UP first	COO + branch ops	Docket printed → OTC: Gujarat 9.9 → 1 d; docket → release > 5 d 44% → 10%	A week off a quarter of West-2 files
4	Tranche queue on the SUD dashboard: stage, days since last tranche, last site visit, RM named, P1–P5	COO + CBO	Idle > 90 d ₹37 Cr → ₹15 Cr	Interest on ₹99 Cr
5	e-Sign and e-NACH mandatory at docket; penny-drop on payee at docket; docket checklist auto-built from sanction conditions and report statuses	CTO + COO	Ops ADRs per file 1.15 → 0.3; HO OTC queries 1,955/yr → 500	Ops capacity; NACH live before first EMI
6	Right-first-time sanction: terms confirmed with the customer before the letter	CRO + CBO	Re-sanction 46% → 25%	~6 days off half the book
7	Payment signatories to Finance/Ops; two payment batches (13:00, 18:00); insurance booking generator with LOS validation	CFO + COO	Releases after 18h 29% → 15%; booking match 100%	Control + a day on a third of releases
8	Technical: 3-day vendor SLA, internal valuation capped, boundary-mismatch and no-plan cases flagged to Collateral Radar	CRO + Technical head	Initiation → report p90 17 → 7 d	Late technical rejects halved
9	Docket Movement Register and Document Return Register as LOS stages — actor, timestamp and discrepancy code on every row; CERSAI creation stamped with asset ID	COO + CTO	Payment → file in storage measurable; closure → documents in customer's hand measurable, target 30 days	Custody control; regulatory clock
10	Data: renames, timestamps, full workflow export, MIS on all logins	CTO	Stages 8–14 measurable by December	Everything above becomes a dashboard

Method and caveats

Event log: 1 Aug 2025 – 14 Sep 2026, 954,641 events, elapsed times recomputed from timestamps; used to confirm every post-sanction gap and to name the OTC compliance and revocation stages.
Windows: post-sanction TATs on real files sanctioned 1 Apr 2025 – 30 Jun 2026 (mature); PDD stock as of 9 Sep 2026; tranche stock and book as of 31 Aug 2026; handover and ADR queries from Apr 2025; fee ledger 1–11 Sep 2026; insurer bookings 8–11 Sep 2026; digital MIS on 550 QC-approved loans since April 2026.
Exclusions: nine digital lead feeds and two bulk lead-initiators (as in the front-end report). Feed files rarely reach sanction, so the effect here is small.
Stage order is the median rank of each timestamp across files carrying all of them; gaps are between consecutive stages in that order. "Docket printed" precedes "Ops done" by about a day and is the field the dump labels cheque printing.
Sub-stage attribution uses the "done by" name on the first occurrence; second-tranche repeats are excluded from stage TATs and counted in Section 5.
PDD age is days from first disbursal to the report date; the report's own Ageing field is populated only on cheque-handover items and contains outliers, so it was not used.
Scope ends at first EMI. Repayment and collections are a separate study; the fee ledger was used only for processing-fee waivers.
Query typing is by keyword on free text; a query can match more than one type. Resolution times: handover report in hours (timestamps present); ADR report in days (dates only).
Not measured: CERSAI, NACH activation, UTR/handover per tranche, policy issuance, welcome call — Section 12.