Shangjian Series C: Ant Group Bets on AI Financial Agents

Shangjian Technology has completed its Series C funding round, led by Ant Group, with participation from BONC and Qiaobei Capital, according to PANews. The funding amount was not disclosed. After the financing, the company and Ant Group plan to cooperate on AI financial agent research and development and on training financial vertical models.

The Shangjian Series C is the latest signal that China's financial giants are no longer just building agents in-house — they are buying into the startups that make agents investable. It arrives on a day when agentic finance is already dominating headlines: the same morning, Today AI announced a $200M+ personal-agent raise backed by Ant and Tencent, as covered on this site, putting two Ant-backed agent raises on the board within hours of each other.

The deal and what comes next

Per PANews, the Shangjian Series C round was led by Ant Group, joined by BONC and Qiaobei Capital. No valuation or round size was announced — a common pattern in China-side strategic rounds where the partnership matters more than the headline number. The stated post-investment agenda is telling: cooperation on AI financial agent R&D and financial vertical model training. That framing suggests Ant wants access to Shangjian's models and agent infrastructure as much as it wants a seat on the cap table.

Vertical-model training is the quiet battleground of the current cycle. General-purpose models can draft a research note, but investment-grade output requires models tuned on financial data, compliance constraints, and institutional workflows. By pairing its capital and transaction ecosystem with Shangjian's domain models, Ant is buying a shortcut into the research desks of thousands of institutions — a distribution channel no foundation-model lab can replicate on its own.

AlphaClaw and the institutional money agent

Shangjian Technology was founded in 2017 by Fei Binjie, according to PANews. Its core product, AlphaEngine, has served more than 100,000 institutional investors across over 9,000 institutions — an installed base that gives any new agent product instant reach into the buy side.

The company released its financial investment research agent, AlphaClaw, in March 2026, and in September it launched an AlphaEngine institutional edition built on a local-first architecture. The local-first detail matters: regulated institutions that cannot ship proprietary data to public clouds can still run the agent on their own infrastructure. That design choice turns a compliance constraint into a distribution advantage, and it is a pattern worth watching across every vertical-agent category — the winners will be the agents that run where the data already lives.

Investment research is an unusually good fit for agentic systems. The job is already a loop: gather filings and filings-adjacent data, build models, cross-check claims against numbers, and write it up. An agent that can run that loop around the clock, with every claim traceable to a line item, is not a demo — it is a workflow replacement. Shangjian's pitch, backed by nine years of institutional tooling, is that AlphaClaw is already that product.

Ant Group's agentic build-out

Ant Group has been assembling the agent stack at every layer, and the Shangjian Series C plugs a gap in the middle. In July, Ant International closed a $1.2 billion Series A, backed by Ant Group and Alibaba, to fund cross-border payment rails and autonomous payment agents, as reported by Startup Fortune. The same company introduced an Agentic Mobile Protocol in April, a framework that lets merchants, AI platforms, and digital wallets plug into agent-initiated payments — with a network of over 150 million merchants and more than 2 billion user accounts behind it.

At the consumer layer, Ant overhauled Alipay around an AI agent named Ah Bao that can perform roughly 10,000 functions, from booking rides to buying mutual funds, according to CKGSB Knowledge. At the enterprise layer, Ant's Agentar 2.0 platform debuted at the World AI Conference with 200 premade digital-expert templates for deploying autonomous AI workers. CKGSB's analysis frames the strategy plainly: while competitors build agents that talk, Ant is building agents that transact, anchored by a billion-user super-app, tens of millions of merchants, licensed financial infrastructure, and two decades of risk-control data.

Read together, the pattern is a full-stack agent economy: consumer agents at the Alipay layer, enterprise agents at the Agentar layer, payment rails at the Ant International layer, and now — with the Shangjian Series C — investment-research agents feeding the institutions that sit on top of all of it. Each layer makes the others more valuable, and each one is defended by data moats that open models cannot cross without partnerships like this one.

What it means for the agent economy

The Shangjian Series C lands the same week as a string of agent-governance and agent-payment stories — including Rhapsody Guardian's clinical-duty launch, also covered on this site — and together they sketch the industry's 2026 playbook: agents are moving from consumer chat into regulated, money-touching workflows, and the capital is following them there. Investment research, with its auditable data trails and institutional budgets, may be the first vertical where agents cross from pilot to procurement at scale.

The open question is what comes after cooperation becomes consolidation. Ant's history with portfolio companies tends toward deep integration rather than passive holding. If AlphaClaw ends up wired into Ant's financial vertical models and, eventually, its transaction rails, the line between research agent and execution agent gets very thin — and that is exactly the territory the rest of the agent industry is racing to claim.