A naming tradition deserves careful research.

In many Indian families, choosing a baby's name begins with a Nakshatra, a Pada or a Rashi — and quickly becomes a family conversation. A birth detail is shared, a starting sound is suggested, and names arrive from relatives, priests, books and WhatsApp groups, with meanings copied from one site to the next.

Before long it's hard to tell which suggestions fit the family's tradition, which meanings are reliable, and where any of it came from. TaraGrid brings the calculation, the names and the supporting evidence into one private report, so you spend less time sorting contradictory results and more time choosing together.

What TaraGrid does

TaraGrid reads two related naming traditions from one birth calculation. The Nakshatra + Pada reading gives a focused starting sound from the Moon's position at the exact time and place of birth — 5 names. The Rashi reading gives a broader set of traditional sounds from the same calculation — 10 names.

Names come from a reviewed corpus of 1,200+ classical Sanskrit names plus clearly labelled modern-origin options. Each carries what's actually known about it: Devanagari spelling, pronunciation, meaning, linguistic root or origin, and a source note — and, where the name is attested in a primary text, the relevant verse. This is naming research — not a horoscope or prediction service.

Our discipline

We would rather show a smaller set of defensible names than a long list of uncertain ones. Four rules hold that line:

Evidence before volume. A classical name appears only if a Sanskrit dictionary or a primary text supports it. Modern-origin names are labelled as such.

Sources, not embellishment. Where a verse is shown, the Sanskrit is reproduced verbatim from the cited text — never reworded to persuade — and not every dictionary-attested name has a verse, so the report never pretends otherwise.

Complete, never filler. Every reading is full — 5 nakshatra names and 10 rashi names. When an exact Pada is sparse, the remaining names come from the same-sound family within the Nakshatra — never from weaker entries or invented names.

Software ranks; it never invents. We use language models in our pipeline only to rank and present names from the reviewed corpus — never to create a name, meaning or citation.

How it works

The product runs on two halves, kept strictly separate. The first is astronomy: from the birth date, local time and place, we compute the Moon's sidereal position in code using astronomy-engine and the Lahiri (Chitrapaksha) ayanamsa. The output is a single Nakshatra, Pada and Rashi and the sounds they carry — deterministic and reproducible. The methodology is published so it can be checked.

The second half is the name corpus: 1,200+ classical Sanskrit names, each verified against Sanskrit dictionaries and cross-referenced with primary sources, alongside clearly labelled modern-origin options. The two halves meet only where the reading's starting sounds select names from the corpus. The astronomy never touches the corpus, and the corpus never touches the astronomy.

TaraGrid organises the research; your language, elders and traditions carry the final choice. We don't expect software to have the final word on a child's name — we expect it to show its work.

Built in India

TaraGrid is an independent baby-naming research service operated by Sukriti Consultants in Gurugram, Haryana. It is not affiliated with any religious institution, astrology service or priestly panel. The corpus is curated by hand and keeps improving as stronger references and corrections come in.

We don't sell birth data — to anyone, ever. Birth and contact details and generated documents are deleted automatically within a retention window of up to 12 months (sooner on request — hello@taragrid.com); the full data policy is in our privacy policy. City and timezone data is drawn from the GeoNames geographical database (CC BY 4.0).

If you spot an inaccurate meaning, a missing citation or a name our corpus should include, write to hello@taragrid.com — we read every email and update the corpus when new evidence lands.