Why deterministic chart context matters for astrology AI
On separating calculation from generation — and why grounding matters more than scale.
Computational astrology infrastructure and practitioner-guided intelligence systems. Free API access is open now. A domain-adapted prediction model is in training.
Built from deterministic astrological calculations, anonymized prediction corpora, and experienced practitioner review.
Generate charts, Panchang data, Dashas, transits, matching inputs, and structured astrological outputs through a developer-friendly API layer.
Rasi, Navamsa, and divisional charts with deterministic planetary positions.
Tithi, vara, nakshatra, yoga, karana, and sunrise/sunset for any location.
Sidereal positions, retrogrades, declinations, and nakshatra padas.
Mahadasha, antardasha, and pratyantardasha periods with precise transitions.
Current and forward planetary transits relative to a natal chart.
Structured inputs for Ashtakoot, Dashakoot, and other matching frameworks.
We are training and evaluating a domain-adapted model on a large historical prediction corpus, supported by review from senior traditional practitioners. The system is being designed for structured reasoning, empathetic language, and tradition-consistent interpretation.
Our work separates deterministic chart computation from interpretive generation. Outputs are evaluated for chart consistency, specificity, empathy, contradiction rate, and safety boundaries.
Generate a token, send a single request, receive structured chart output. The same endpoints power our internal evaluation harness.
curl https://api.astroresearchlabs.com/v1/chart \
-H "Authorization: Bearer arl_test_your_key" \
-H "Content-Type: application/json" \
-d '{
"date": "1994-08-21",
"time": "14:35:00",
"timezone": "Asia/Kolkata",
"latitude": 28.6139,
"longitude": 77.2090,
"ayanamsa": "lahiri"
}'Short technical writeups on the choices behind the calculation layer, the evaluation framework, and the practitioner review loop.
On separating calculation from generation — and why grounding matters more than scale.
A pragmatic evaluation framework: consistency, specificity, empathy, contradiction.
Notes from designing endpoints around Panchang, Dasha, and transit semantics.