Pefindo Product 132

Endpoint: tmb-eagle-api.idscoresystem.id
Product ID: 132
Inquiry reason: 15
Member: MILIKIRUMAH · 372
Payload size: 7.4 – 344 KB
Latency: ~5.2 s

01 — OrientationThe shape of it

Product 132 is a substantially reduced tier against the P1 structure spec. The spec describes ~170 debtor fields and ~90 per facility. What arrives is exactly 40 and exactly 13 — identical on every one of the 17 subjects tested, whether they had one facility or 705.

debitur fields

facility fields

4,162

facilities seen

253

distinct lenders

monthly scores

collateral rows

Evidence base

Every claim here is measured, not inferred from the spec. 17 subjects drawn from Pefindo's own staging test lists — including a list explicitly titled "Debitur dengan agunan personal" (debtors with personal collateral) and one titled "fasilitas banyak" (many facilities). Counts in the n / 17 column say how many subjects support each finding.

02 — Call sequenceFour calls, about five seconds

GET /api/v1/getToken

Basic auth. JWT at data.token, valid 3600 s.

0.22 s

POST /api/v1/product/search

Returns inquiry_id plus candidates carrying id_pefindo.

0.07 s

POST /api/v1/product/generateReport

You mint the event_id UUID. Reuse is rejected with code 35.

0.04 s

GET /api/v1/product/getReport/eventId/{uuid}

Returns code 32 while scoring runs. Poll until 01.

5.11 s

One poll cycle was enough for all 17 subjects, including the 705-facility one — report build time does not scale visibly with portfolio size.

Request body

POST /api/v1/product/search
{
  "type": "PERSONAL",
  "product_id": 132,
  "inquiry_reason": 15,
  "reference_code": "your-correlation-id",
  "params": [
    {
      "id_type": "KTP",
      "id_no": "3529116903740002",
      "name": "TRI MURYATUN",
      "date_of_birth": "1982-12-17"
    }
  ]
}

Entitlement

Product 132 with reason 15 is what MILIKIRUMAH is provisioned for. Documented products 123 and 148 return "Akses tidak diberikan"; reasons 12/13/14 from appendix 3.4.3 return "Inquiry Reason Tidak Valid" — Eagle does not recognise them at all.

03 — EnvelopeWhere each section lives, and how often

The important column is n / 17. Several sections are not simply "present" or "absent" — they appear for some subjects and not others, with no relationship to portfolio size.

Section	Type	n / 17	Row counts seen	Verdict
report.header	object	17	3 fields	Always
report.debitur	object	17	40 fields	Always
report.fasilitas	array	17	1 – 705	Always
scoring (top level)	array	17	13	Always
report.riwayat_identitas_debitur	array	14	5, 9, 10, 23	Usually
report.summary_riwayat_debitur	array	6	13, 24	Sometimes
report.permintaan_data	array	0	—	Never
report.summary_permintaan_data	array	0	—	Never
report.pengurus	array	0	—	Never
report.other_data	array	0	—	Never
fasilitas[].riwayat_fasilitas	array	0	0 / 4,162 fac.	Never
fasilitas[].agunan	array	0	0 / 4,162 fac.	Never
fasilitas[].penjamin	array	0	0 / 4,162 fac.	Never
CurrentContract · otherDataSource	null	0	null	Null
InquiriesAnalysis · PastDueInformation	null	0	null	Null
RepaymentInformation	null	0	null	Null

Revised in this edition

An earlier draft, based on two subjects, recorded summary_riwayat_debitur as reliably present. It is not — it appears for 6 of 17 subjects. Critically, absence does not track portfolio size: subjects with 705, 677 and 524 facilities returned zero months of history, while others with comparable portfolios returned 24. Nothing in the payload predicts which you will get.

Trap 1 — scoring is not inside report

scoring[] is a sibling of report, not a child of it, though the API doc's worked example nests it. Reading report.scoring returned an empty result on all 17 subjects — no error, no empty array, just a missing key. Read payload.scoring.

Collateral: the decisive test

Six of the 17 subjects came from a Pefindo staging list titled "Debitur dengan agunan personal" — debtors with personal collateral. All six returned agunan absent on every facility. Across all 17 subjects and all 4,162 facilities, the count of facilities carrying agunan, riwayat_fasilitas, or penjamin is zero. These blocks are withheld at this tier, not merely sparse.

04 — report.debiturForty fields, invariant

The same 40 keys on all 17 subjects. Every id_* field is a code requiring a lookup table. Note the absence of the sector-prefixed aggregates (bank_*, mf_*, p2p_*…) the spec describes — no sector breakdown exists at this tier.

Field	Type	Meaning	Example
nama_lengkap_debitur	string	Full name	TRI MURYATUN
nama_sesuai_identitas	string	Name as on ID document	TRI MURYATUN
nomor_identitas	string	KTP number	3529116903740002
npwp	string	Tax number	3529116903740002
id_debitur_golden_record	int	Pefindo ID — the stable subject key	102171282000021
tanggal_lahir	string	Date of birth	1982-12-17
tempat_lahir	string	Place of birth	SUMENEP
nama_gadis_ibu_kandung	string	Mother's maiden name	HJ ARIFAH
id_jenis_kelamin	string	Gender · L / P	P
id_status_perkawinan	string	Marital status	1
id_status_gelar	string	Education level	00
id_pekerjaan	string	Occupation code	099
tempat_bekerja	string	Employer	SUMENEP
alamat_tempat_bekerja	string	Employer address	SUMENEP
alamat_debitur	string	Address	DUSUN KACANGAN…
kecamatan	string	District	PAKAMBAN LAOK
kelurahan	string	Sub-district	PAKAMBAN LAOK
id_lokasi	string	Location code — use instead of city	69465
id_kabupaten_kota	string	City — "0000" on every subject tested	0000
id_negara	string	Country	ID
email	string	Email	TRIMURYATUN@…
telepon	string	Phone	081999026024
telepon_seluler	string	Mobile — trailing spaces	"081999026024 "
id_tipe_debitur	int	Debtor type · 1 = individual	1
id_golongan_debitur	string	Debtor group	S14
id_sektor_ekonomi	string	Economic sector	X-013000
jml_aktif_fasilitas	int	Open facilities	8
jml_tutup_fasilitas	int	Closed facilities	137
jml_fasilitas_tertunggak	int	Delinquent facilities	0
jml_fasilitas_surat_berharga	int	Securities facilities	0
jml_plafon	int	Total credit limit — opaque, Trap 3	156,893,774
jml_saldo_terutang	float	Total outstanding — opaque, Trap 3	361,778,471
jml_tunggakan	float	Total overdue	20,000.0
tunggakan_terburuk	float	Worst overdue amount	20,000.0
jml_hari_tunggakan	int	Total days past due	180
jml_hari_tunggakan_terburuk	int	Worst DPD — unreliable, Trap 2	0
kolektabilitas_terburuk	int	Worst collectibility — unreliable, Trap 2	2
tgl_buka_fasilitas_terakhir	string	Last facility opened	2023-11-30
tgl_tutup_fasilitas_terakhir	string	Last facility closed	2024-04-05
tgl_tunggakan_terakhir	null	Last overdue date — null on every subject	null

05 — report.fasilitas[]Thirteen fields, up to 705 rows

Identical field set on all 4,162 facilities observed. No nama_ljk — the lender resolves from id_pelapor. No id_kolektabilitas — there is no per-facility collectibility at this tier.

Absent versus spec: nama_ljk, id_kolektabilitas, saldo_terutang, id_jenis_pelapor, kantor_cabang, suku_bunga_atau_imbalan, tunggakan_pokok, id_valuta, nomor_akad_*, and ~75 others.
Field	Type	Meaning	Observed range
id_pelapor	int	Lender id — signed 32-bit, often negative	253 distinct values
id_jenis_kredit	string	Credit type	10 distinct codes
id_kondisi	int	Facility condition	1,3,4,5,6,8,9,10,13
fase_fasilitas	int	1 = open, 2 = closed	1 · 2
plafon	float/int	Credit limit	0 – 170,000,000
baki_debet	float/int	Principal outstanding	0 – …
nominal_tunggakan	float	Overdue amount	0 – …
tunggakan_terburuk	float/int	Worst overdue amount	0 – …
jumlah_hari_tunggakan	int	Current days past due	mostly 0
jml_hari_tunggakan_terburuk	int	Worst days past due	0 – 23
tanggal_awal_kredit_atau_pembiayaan	string	Start date	2011 – 2024
tanggal_akhir	string	End date	2011 – 2024
tahun_bulan_data	string	Reporting period of this row	2017-03 – 2024-05

Facility condition across all 4,162 facilities

85 of 4,162 facilities (2.0%) carry a write-off condition (4, 5, 6 or 13).
id_kondisi	Meaning	Count	Signal
3	Lunas — settled	3,822	Good
1	Fasilitas Aktif — active	152	Neutral
4	Dihapusbukukan — written off	44	Adverse
13	Lunas dengan diskon — settled at a discount	40	Adverse
9	Dialihkan ke fasilitas lain	36	Watch
8	Dialihkan atau dijual ke pelapor lain	35	Watch
10	Dialihkan ke pihak lain non-pelapor	25	Watch
6	Lunas karena pengambilalihan agunan — collateral seized	8	Adverse
5	Hapus tagih — debt forgiven	1	Adverse

Use condition, not DPD, as the adverse signal

Facility-level jumlah_hari_tunggakan was 0 on almost every facility tested, and jml_hari_tunggakan_terburuk never exceeded 23 days. Yet one 705-facility subject carries 14 written-off facilities. Delinquency is simply not expressed through the DPD fields at this tier — id_kondisi is where it shows up. Build adverse rules on condition codes.

Lender resolution

All 253 distinct id_pelapor values resolved against the reporter registry extracted from Pefindo's structure workbook — a 100% hit rate across 4,162 facilities. Sector must then be inferred from the lender's name: that yields bank 60.4%, multifinance 17.6%, and 22.0% unclassified, since names like "PT Woka Internasional" carry no sector marker.

06 — summary_riwayat_debitur[]The history block — when you get it

Three fields per month at portfolio level: no amounts, no per-lender attribution. Present for only 6 of 17 subjects, at either 13 or 24 rows.

Field	Type	Meaning
tahun_bulan_data	string	Month-end period, newest first
kolektabilitas_terburuk	string	Worst collectibility that month · 1 Lancar → 5 Macet
jml_hari_tunggakan_terburuk	int	Worst days past due that month · seen up to 2,653

Trap 2 — the debtor aggregates understate the history

debitur.kolektabilitas_terburuk was null on 6 of 17 subjects, and on 4 of the 6 subjects that had history it reported a better grade than the history contains. One subject reads 2 (Dalam Perhatian Khusus) against a history holding ten months at 5 (Macet) and 266 days past due; another reads null against a history reaching Macet.

Taken together the field is unusable on 10 of 17 subjects. Always compute worst-case across summary_riwayat_debitur where it exists, and fall back to id_kondisi write-off flags where it does not.

Worked example — one subject's 24 months

Ten adverse months of 24. The debtor aggregate for this subject reports collectibility 2 and worst DPD 0.
Period	Collectibility	DPD	Status
2024-05 → 2023-05	1 · Lancar	0	13 months clean
2023-04	2 · Dalam Perhatian Khusus	20	Watch
2023-03 → 2022-12	5 · Macet	180	Default
2022-11 → 2022-10	5 · Macet	266	Default
2022-09 → 2022-06	5 · Macet	264	Default

07 — scoring[]Thirteen monthly scores, always present

Exactly 13 entries on every subject, oldest first, ending with a score dated the day of the request. Present universally — but calculable on only 10 of 17.

200–624

real score range

10 / 17

scorable

7 / 17

unscorable

risk grades seen

Field	Type	Meaning	Example
period	string	Scoring period	2026-08-18
score	int	Score — 9999 is a sentinel	624
pod	float/int	Probability of default, %	10.23
risk_grade	string	Grade code	RRD1
risk_grade_desc	string	Grade description	Risiko Tinggi
risk_desc	string	Risk description	Risiko Tinggi
reason_code	array	Up to 5 reason codes	["NSS1","MSD6"]
reason_desc	array	Matching descriptions	…
id_pefindo	string	Subject key	102171282000021

Grades observed across all series, best to worst: RRC1, RRC2, RRC3, RRD1, RRD2, RRD3, RRE1, RRE3, plus XX for unscorable. Response language follows the language parameter on generateReport: 01 Bahasa, 02 English.

Trap 3 — 9999 is not a score

Seven of 17 subjects returned score: 9999, risk_grade: "XX", pod: 100, reason ND01 — "score cannot be calculated, at least 2 months of payment information required in the last 24 months". That is a sentinel meaning unscorable, not a score of 9999 or a 100% default probability. At 41% of subjects this is a common path, not an edge case. Gate on score == 9999 || risk_grade == "XX".

Scorability does not follow portfolio size: one single-facility subject scored 582, another single-facility subject was unscorable.

08 — Implementation notesEverything else that bites

Trap 4 — the money aggregates do not reconcile

An earlier draft reported that jml_saldo_terutang ties exactly to sum(baki_debet) + sum(nominal_tunggakan). Across 15 subjects it ties on only 5; on the rest it is understated by 5–20%. jml_plafon is worse: it matches the sum of all facility limits on 3 subjects, matches the open-facility sum on 5, and matches neither on 7 — one subject reports 746 M against a facility total of 8.18 bn.

Treat both as opaque bureau-computed figures. Do not derive them, do not validate against them, and do not compute utilisation from their ratio.

Parsing

tgl_permintaan is a Java ZonedDateTime: 2026-08-18T11:06:47.187796165+07:00[Asia/Jakarta]. Strip the bracketed zone id before parsing.
telepon_seluler carries trailing spaces. Trim all strings.
id_pelapor values are signed 32-bit and frequently negative (−2079932240). Do not assume unsigned.
Money fields arrive as float or int inconsistently for the same field across records. Coerce.
id_kabupaten_kota was "0000" on every subject tested — city is effectively unavailable; use id_lokasi.
tgl_tunggakan_terakhir was null on every subject tested.
Payloads reach 344 KB and 705 facilities. Size limits and timeouts should assume an order of magnitude more than a typical file.

Transport

Business failures arrive with HTTP 200 (31 not found, 32 still processing), while others arrive as HTTP 400 or 500 with code: 99. Always read code; HTTP status alone is not a success signal.
The token is at data.token. Pefindo's own Postman script reads data.access_token. Accept either.
event_id must be unique per request and is the only handle for re-fetching a report or its PDF. Persist it with the application record.

09 — AssessmentWhat this supports, and what it doesn't

As a fallback bureau source, product 132 answers portfolio-level questions and facility-level ones not at all. Two of its most useful signals are also conditional on data that arrives only sometimes.

Underwriting question	Supported	Source · coverage
Does this person have credit history?	Yes	fasilitas[] · 17/17
What is their credit score and PD?	Often	scoring[] · 10/17 calculable
How much do they owe in total?	Approx	aggregates · do not reconcile
How many lenders, and who?	Yes	id_pelapor · 253/253 resolved
Has anything been written off?	Yes	id_kondisi 4/5/6/13
Have they defaulted in 24 months, how badly?	Sometimes	summary_riwayat_debitur · 6/17
Current delinquency and days past due?	No	DPD fields ~always 0
Which facility went bad, and with whom?	No	needs riwayat_fasilitas
Per-facility collectibility?	No	no id_kolektabilitas
Month-by-month balance and arrears trend?	No	needs riwayat_fasilitas
Credit-hungry — recent inquiry velocity?	No	permintaan_data · 0/17
What collateral secures the exposure?	No	agunan · 0/4,162 facilities
Exposure split by lender sector?	Partial	inferred · 22% unclassified

Open with Pefindo

Are riwayat_fasilitas, agunan and permintaan_data available on a higher tier, or withheld from all API products? The "debitur dengan agunan personal" subjects returned no collateral at all.
Why does summary_riwayat_debitur arrive for some subjects and not others, with no relation to portfolio size?
Why do debitur.kolektabilitas_terburuk and jml_hari_tunggakan_terburuk disagree with the history they summarise, and why is the former frequently null?
How are jml_plafon and jml_saldo_terutang computed, given they match neither the full nor the open facility totals?
Can product 123 (IdScore+) be enabled for comparison?

Rev. 2 · compiled from 17 live captures against tmb-eagle-api.idscoresystem.id, 18 Aug 2026 · member MILIKIRUMAH · product 132 · inquiry reason 15 · 4,162 facilities · 253 distinct lenders · staging test data