AWS ProServe Unveils ‘Artham’, India’s First Small Language Model for Capital Markets, Built by Raise Financial Services
Dr Swami Sivasubramanian, Vice President, AWS Agentic AI, delivers a keynote address at AWS re:Invent 2025.
New Delhi, December 4: In a major milestone for India’s fintech and AI ecosystem, Amazon Web Services (AWS) showcased Artham — a 7-billion-parameter Small Language Model (SLM) trained exclusively on Indian financial and capital markets — at its flagship AWS re:Invent 2025 conference in Las Vegas. Built by Mumbai-based Raise Financial Services, Artham becomes the country’s first domain-specific SLM designed for investors, traders and financial professionals.
Artham was unveiled during a keynote session by Dr Swami Sivasubramanian, Vice President, AWS Agentic AI. AWS highlighted Raise as a standout example of customers building deep, domain-oriented AI products on its platform — a rare recognition of an Indian startup featured prominently on the global stage. The announcement follows an earlier keynote by AWS CEO Matt Garman, who named Raise’s trading platform Dhan among the early adopters of AWS DevOps Agent.
Positioned as “The Financial Brain, Built for India,” Artham is engineered to understand India’s financial vocabulary, regulatory architecture and market behaviour — a capability global general-purpose AI models lack. Developed over nine months by Raise AI, the model is trained on a unique 70:30 blend of public and proprietary Indian financial datasets, including company disclosures, regulatory filings, and macroeconomic information. Over time, the proprietary share is expected to grow as Artham deepens its coverage.
The model is fully hosted within India to ensure data sovereignty and sectoral compliance. Artham also features native tool-calling, enabling secure access to real-time market data and analytics. This capability allows Artham to generate source-backed insights — a key requirement for regulated financial platforms. It already powers Raise’s products, including fuzz (askfuzz.ai), the AI-enabled financial research assistant, and ScanX (scanx.trade), a markets intelligence platform. Portions of news generation and research insights on Dhan and ScanX are now driven by Artham.
Pravin Jadhav, Founder & CEO of Raise Financial Services, said Artham is built as an India-first financial AI layer designed for precision, explainability and compliance. Co-founder and CTO Alok Pandey added that a lean internal team at Raise AI was able to move from experimentation to production within months, aided by AWS infrastructure.
Raise emphasised strong guardrails around the model’s deployment — Artham provides information and educational insights, not investment recommendations. The company clarified that users remain responsible for investment decisions.
The launch marks a key moment for India’s financial technology landscape, with a home-grown AI system now powering real-time market analysis, investor education and research workflows. Raise said Artham will serve as a foundational capability for new research, learning and analytics tools tailored to the needs of India’s maturing financial markets.
Raise Financial Services, founded in 2021, operates a suite of platforms including Dhan, fuzz, ScanX, Upsurge and Filter Coffee, backed by leading global investors such as MUFG, Mirae Asset and BEENEXT.