Janma · जन्म · Birth Reconstruction · Pro v2

Janma Finder Enginean evidence engine for birth-date discovery & birth-time rectification

Not a kundali calculator. This tool ranks candidate birth dates using panchāṅga clues, Moon nakṣatra and Vimśottarī dāśā sequence, then rectifies birth time against your dated life events, Lagna, body and family pattern — and reports an honest confidence with every failed match shown.

You know the date and place; the birth time is approximate. The engine sweeps your time window and rectifies to the second.

1 Birth Data & Place

All computation is sidereal (Lahiri). Coordinates and UTC offset drive the Lagna and dāśā balance — get the longitude right.

Default: Jaipur, India (+5.5). Western longitudes and offsets are negative.

1B Date Discovery Clues

Used to rank candidate dates before time rectification. Crucial for Mode B/C; harmless in Mode A. Family memories of festival, tithi, weekday or nakṣatra are often more reliable than a school-record date.

2 Birth Circumstances

Low weight, but good for eliminating weak candidates — a C-section points to Mars/Ketu/8th, water nearby to Moon/Cancer/Pisces.

3 Body, Temperament & Lived Pattern

A "lived-pattern signature." These convert observed traits into Lagna / Moon / planetary signatures and act as a moderate tie-breaker (heavier in Mode C). They never override dated events. Rate 0 (not at all) → 5 (strongly).

Health items are pattern signatures, not medical diagnosis. Surgery/illness with a date should also be entered as a dated event below — that's where they carry real weight.

4 Parents & Family

Parents validate the 4th/9th axis and D12. A parent's death date is one of the strongest anchors there is — enter it as a dated event in the timeline; the toggles here add supporting natal-pattern checks.

5 Life-Event Timeline the engine

This is what actually rectifies the chart. Give 6+ dated events; mark the heavy ones (deaths, marriage, surgery) honestly. Accuracy now matters: a "month"-level date is weighted ~0.65, "year" ~0.35 of an exact date.

Event type
Date
Accuracy
Weight

Calibration Self-Test

Confidence is meaningless unless the engine can recover a time it doesn't already know — and unless each scoring layer actually helps. This test hides a true time, generates consistent events, recovers it, then ablates: it removes each layer in turn to see whether accuracy gets worse (the layer earns its place) or doesn't (the layer is decoration). It checks internal consistency, not real-world accuracy (that needs a verified-chart database, which isn't built here).

Most Probable Birth Moment
acceptable range —
Birth date
Lagna
Lagna degree
Moon nakṣatra
Tithi
Dāśā at birth
Events matched

Rāśi Chart (rectified)

North-Indian diamond · houses fixed, signs rotate · Lagna in house 1

Evidence layers

Weighted blend behind this moment. Larger bars pulled harder toward the chosen time/date.

6 Event-Matching Ledger

Every event scored on natal promise, dāśā at the event date, transit trigger, and KP cusp sub-lord (real Placidus). Failures are shown, never hidden.

7 Contradiction Audit

Where a heavy event the native lived through is not supported by the chosen chart. Each contradiction lowers confidence.

8 Candidate Time Windows

Top scoring candidate moments. A lone spike with neighbours far below is a warning sign, flagged automatically.

Honesty notice. This is a prototype reasoning engine, not a production pipeline. The astronomy has been tightened: Sun & Moon are accurate to ~an arc-minute, the Moon is now topocentric (parallax-corrected, which can shift its nakṣatra near a boundary), positions use nutation and a true (apparent) ayanāṁśa, the five planets use a proper geocentric reduction good to ~1°, and house cusps are now real Placidus (validated as monotonic at all latitudes) rather than equal-house — so the KP cusp sub-lord layer is genuine, not a proxy. What still falls short of research-grade: planets are ~1° not arc-second (needs Swiss Ephemeris / JPL), the KP ayanāṁśa is modeled as a small offset from Lahiri, divisional scoring is a lightweight proxy not full varga computation, timezone/DST is taken from your manual UTC offset rather than a historical database, and behaviour/health/family layers are supportive pattern-matching (health is never a medical diagnosis). Use the Calibration Self-Test above to see the engine recover a hidden time; a separate deployable Swiss-Ephemeris service (Python/FastAPI) accompanies this file for the precision upgrade. The output is a confidence-ranked hypothesis, not a birth certificate. Astrology is not predictive science.