We believe every Nepali investor deserves access to institutional-grade AI trading intelligence — not just hedge funds.
To democratize AI-powered financial intelligence for Nepali investors — delivering institutional-grade NEPSE analysis, predictive signals, and automated tools that were previously accessible only to large institutions and professional traders.
To become Nepal's most trusted AI fintech platform — transforming how NEPSE is analyzed, understood, and traded through cutting-edge machine learning, real-time data infrastructure, and intuitive user experiences.
Generic trading platforms don't understand NEPSE's unique dynamics. We do.
Every algorithm, every data model, every tool is built specifically for Nepal's stock exchange — not adapted from foreign markets.
Our LSTM models are trained on years of NEPSE historical data and continuously retrained on fresh market data — not backtested toy models.
Your portfolio data, trading patterns, and personal information are encrypted and never sold to third parties. Your data is yours.
Our data pipeline processes NEPSE tick data in milliseconds — signals arrive before the crowd reacts, giving you a measurable edge.
We don't just give you signals — we explain the reasoning. Every recommendation comes with educational context to build your own skills.
Our support team is based in Kathmandu, speaks Nepali, and understands the local market context. Real help, not chatbots.
ASHVA Tech was founded in Kathmandu with a mission to bring AI-powered analysis to Nepali investors.
First production LSTM model trained on 5 years of NEPSE data. Achieved 82% directional accuracy in backtests.
Launched the web dashboard with 200+ beta users. Added real-time sector heatmaps and portfolio analytics.
Crossed 500 active subscribers. Launched ASHVA Bot with backtesting engine and Telegram integration.
Built on battle-tested ML frameworks and modern cloud infrastructure.
Core ML pipeline with LSTM, Transformer, and ensemble models.
WebSocket-based NEPSE data ingestion with sub-second latency.
Scalable cloud compute for model training and serving.
End-to-end encryption, JWT auth, and SOC2-aligned security practices.