Itoflow Raises $2.5 Million to Build AI Agents for Investment Research and Portfolio Management

Itoflow, a startup building AI agents for professional investment teams, has raised $2.5 million in pre-seed funding as it looks to automate parts of investment research, portfolio monitoring and risk management.

Balderton Capital led the round. The funding also attracted angel investors, including Barney Hussey-Yeo, founder and CEO of personal finance company Cleo.

Founded in 2026, Itoflow is taking aim at a problem that has long challenged smaller investment firms: building the research infrastructure needed to test investment ideas, monitor portfolios and manage risk can require significant amounts of specialist talent, data and engineering.

The startup’s pitch is straightforward. Instead of asking each investment firm to build those systems from scratch, Itoflow wants AI agents to turn the firm’s own investment process into continuously running, governed workflows.

Itoflow wants AI to work around each firm’s investment process

The platform is not designed around a single investment strategy.

Instead, investment teams can describe their approach, risk constraints and review criteria in plain English. Itoflow’s agents then use those instructions to conduct quantitative research, run backtests and monitor portfolios.

The system can also flag material changes and provide supporting evidence for its findings.

That distinction is important.

An investment manager does not necessarily want an AI system imposing an unfamiliar strategy on the portfolio. Instead, Itoflow is trying to make the firm’s existing process easier to scale.

The company says its technology can work across asset classes including equities, bonds, ETFs, commodities and digital assets.

From quantitative research to continuous monitoring

Itoflow’s platform is designed to cover several stages of the investment process.

An analyst can use the system to investigate an idea. The agents can then test the underlying hypothesis through quantitative research and backtesting.

Once a strategy or portfolio is being monitored, the system can continue looking for significant changes.

That creates a different model for investment research.

Instead of relying solely on analysts to repeatedly check portfolios and market developments, an AI agent can operate continuously and bring important changes to the team’s attention.

The aim is not necessarily to replace portfolio managers.

Rather, Itoflow is positioning the technology as an additional research and monitoring layer that allows investment teams to cover more ground without having to expand their research staff at the same rate.

Human approval remains part of the system

Investment decisions come with a much higher level of responsibility than many consumer AI applications.

A mistake in a research assistant can be inconvenient. A mistake in a system monitoring a live portfolio can have direct financial consequences.

Itoflow has therefore built human controls into its platform.

Customers can determine which actions require human approval and set thresholds governing how the agents operate. Institutional clients can also deploy the technology within their own infrastructure, allowing sensitive data and proprietary investment strategies to remain in their existing environments.

That approach reflects one of the biggest challenges facing AI in professional investing: firms want automation, but they also need accountability.

For Itoflow, keeping the investment team in control is therefore part of the product rather than an afterthought.

Itoflow is already running investment-firm pilots

The company is not starting entirely from the laboratory.

Itoflow says it is currently running pilots with hedge funds and family offices and is also in discussions with exchanges, brokerages and financial-data providers.

Outside reporting has provided additional detail on those early deployments.

Tech Funding News reported that Itoflow is running three enterprise pilots, including an asset manager, an ETF provider and a mid-sized hedge fund. The largest engagement is reportedly monitoring roughly $3 billion in assets**.

That figure should be viewed as the size of an asset base being monitored in a pilot, rather than assets managed by Itoflow itself.

The distinction matters because Itoflow is an investment technology provider, not an asset manager.

The founders bring experience from finance and big technology companies

Itoflow was founded by three graduates of the Indian Institute of Technology.

Chief executive and co-founder Aditya Jha spent nine years working as a quantitative researcher at JPMorgan, RBC and Tower Research Capital. At Tower Research Capital, he most recently served as business head for mid-frequency trading.

The wider founding team has also worked on large-scale technology infrastructure at companies including  Apple, Google and Microsoft**, while bringing additional quantitative research experience from JPMorgan.

That combination is central to Itoflow’s pitch.

The company is trying to combine the infrastructure discipline of large technology companies with the quantitative research experience found inside sophisticated trading firms.

Why investment research is a difficult AI problem

Investment research is not simply a matter of asking an AI model which stock looks attractive.

A professional investment process can involve data collection, statistical analysis, backtesting, portfolio construction, risk limits, monitoring and documentation.

The information can also change quickly.

A strategy that worked six months ago may stop working as market conditions change. A relationship between two assets can weaken. A previously useful signal can disappear.

That creates a particularly difficult problem for AI systems.

An agent needs to do more than generate plausible answers. It needs to work with data, follow defined constraints, produce evidence and recognize when the assumptions behind an investment idea no longer hold.

Itoflow’s long-term research goal is to develop agents that can do more of that work themselves while remaining within parameters defined by investment teams.

The long-term ambition is self-improving investment agents

Aditya Jha says the company’s longer-term goal is to build self-improving investment agents**.

The idea is that an agent working with an investment mandate over months or years could learn which approaches continue to work, identify new signals and recognize when a previously successful strategy has lost its edge.

That is a considerably more ambitious goal than today’s research assistant.

A conventional AI research tool responds to requests.

A more autonomous investment agent would continuously observe a mandate, evaluate new information and identify opportunities or risks without waiting for an analyst to ask a question.

Itoflow says that development would still take place within parameters set by each investment team.

Balderton sees an infrastructure gap in investment management

Balderton Capital’s investment thesis centers on the cost and complexity of building institutional-quality research systems.

The venture firm said sophisticated investment organizations can spend years building the systems and research teams required to test ideas, manage risk and respond to market changes.

Itoflow is attempting to make those capabilities available to a broader range of investment firms.

That could be particularly relevant for smaller hedge funds, family offices and asset managers.

Large quantitative firms can afford teams of researchers, engineers and data specialists.

Smaller firms may have strong investment expertise but lack the resources to build the same technology stack.

If Itoflow can successfully package those capabilities into a single platform, the economics of investment research could change.

The startup faces a growing AI investment market

Itoflow is entering an increasingly crowded market.

Financial institutions are experimenting with AI for research, portfolio construction, risk monitoring and client services.

Several startups are also building AI tools for investment professionals.

That makes differentiation important.

Itoflow’s argument is that its agents do not simply provide generic investment information. They are designed to learn and execute **each firm’s own investment process**.

The company also emphasizes governance and human control.

Whether that approach proves sufficiently differentiated will depend on how well the system performs in real investment environments and whether institutional clients are comfortable allowing AI agents to operate continuously around their portfolios.

Funding will support engineering, research and regulatory work

Itoflow plans to use the new $2.5 million to expand its engineering and quantitative research teams.

The company will also invest in product development and the commercial and regulatory work needed as it moves from pilots toward broader adoption.

That spending plan reflects the challenges of selling AI into financial institutions.

Building a useful model is only one part of the process.

Investment firms also need strong security, reliable infrastructure, data controls and clear governance before they can allow an AI system to interact with sensitive investment information.

For Itoflow, the next stage will therefore involve proving both the technology and the operating model.

What Itoflow’s funding means for financial AI

The investment highlights a broader shift in financial technology.

The first wave of generative AI in finance largely focused on assistants that could summarize documents, search information or answer questions.

The next wave is moving toward agentic systems that can perform longer sequences of work.

Investment research is an obvious area for that shift.

A research agent could gather data, test a hypothesis, run a backtest, compare results and prepare evidence for an analyst.

A monitoring agent could then watch the portfolio and flag changes that meet predefined conditions.

The human investment team remains responsible for the decisions, while AI handles more of the repetitive analytical workload.

That is the model Itoflow is trying to build.

The challenge is proving that autonomy can be trusted

The most interesting question for Itoflow is not whether AI can produce investment research.

It is whether investment firms will trust AI to operate continuously around real portfolios.

Markets do not follow office hours.

A significant event can happen overnight. A risk threshold can be breached while a portfolio manager is unavailable. An AI system may also encounter conflicting data or reach a conclusion that requires additional context.

Itoflow’s human-approval controls are designed to address some of those risks.

However, the company will ultimately have to demonstrate that its agents can behave predictably when market conditions become difficult.

That will likely matter more to institutional customers than an impressive product demonstration.

Conclusion

Itoflow’s $2.5 million pre-seed round is a bet that AI agents can make sophisticated investment research and portfolio monitoring more accessible to a wider range of financial firms.

The company’s approach is different from a generic AI investment assistant.

Itoflow wants to encode each firm’s own strategy, risk rules and review process into governed workflows that can operate continuously across markets and asset classes.

The early pilots suggest that institutional investors are willing to test that model. The reported pilot involving roughly $3 billion in monitored assets also gives the startup an early opportunity to prove its technology at meaningful scale.

But the bigger test is still ahead.

Investment firms will need evidence that AI agents can produce reliable research, respect risk constraints and remain accountable when markets move quickly.

If Itoflow can establish that trust, its technology could help smaller investment teams access capabilities that once required large quantitative research and engineering departments.

For now, the company is still in the pilot stage.

But its ambition is clear: turn the investment firm’s own process into an AI system that can research, monitor and learn continuously without taking control away from the people responsible for the portfolio.

Itoflow raises $2.5m for investment research AI agents

AIAI agentsAI investingasset managementFinancial AIfinancial technologyfintechhedge fundsInvestment Researchinvestment technologyItoflowportfolio managementQuantitative FinanceStartup Fundingventure capital