aisot AI Insights Platform
A Swiss platform from ETH Zurich spin-off Aisot Technologies that turns market data into forward-looking quantitative signals, letting portfolio managers, family offices and independent wealth managers build, optimize, backtest and monitor portfolios, then issue them as tradable structured notes.
What is aisot AI Insights Platform?
The aisot AI Insights Platform is a web-based decision-support system for professional investors, published by Aisot Technologies, a spin-off of ETH Zurich based in Zurich. It positions itself as an intelligence layer for portfolio and wealth managers, turning large and complex investment data into AI-driven quantitative signals that inform tactical allocation and risk decisions across the full portfolio management lifecycle: strategy, backtesting, rebalancing, optimization and monitoring. The platform is organized in three parts. The AI Insights Dashboard is the working environment: a portfolio builder to assemble holdings from a chosen asset universe, an optimizer that weighs historical data, market trends and stated preferences to propose allocations, unlimited simulation runs so scenarios can be explored without metering, and a statistical suite covering volatility and risk-adjusted returns. A structuring and feature-request credit is bundled into the subscription. The Custom Feature Suite adds on-demand analytics that clients can request and even vote on: LLM-derived news sentiment, ESG components usable as portfolio constraints, factor investing models, dedicated advanced risk models, dedicated machine learning models, and a backtesting and rebalancing API. The Product Launch Pad closes the chain by converting a strategy into a tradable security, issuing Swiss structured notes intended for institutional and professional investors. Underneath sits aisot's own language model work. Built on the open-source Llama 400B base and fine-tuned quarterly for financial tasks, these models are time-boxed: they only see data that existed at the moment of each forecast, which is how aisot addresses look-ahead bias in backtesting. They are reachable from the platform, or downloadable for integration into a client's own machine learning workflows. Signals are offered on weekly, monthly, quarterly and yearly horizons across Swiss, European, UK, US and Indian markets. Audiences named by the publisher are active funds, family offices, external and independent asset managers, and corporate treasuries. aisot claims up to 90% of time saved on portfolio management tasks, drawdowns reduced by up to 22% and a higher Sharpe ratio; these are vendor figures, published without methodology or test period. The stated philosophy keeps the human in charge: the tool supplies a systematic foundation for decisions that remain the manager's own.
What it does
- Generate forward-looking AI quantitative signals over weekly, monthly, quarterly and yearly horizons
- Build custom portfolios from a chosen asset universe and optimize risk and return under your own constraints
- Run unlimited simulations and backtest strategies before committing capital
- Measure volatility, drawdown and risk-adjusted returns with the built-in statistical tool suite
- Get plain-language explanations of performance, risk and their drivers from the agentic AI layer
- Apply investment restrictions and ESG constraints directly inside the optimizer
- Issue a validated strategy as a tradable Swiss structured note through the Product Launch Pad
When to use aisot AI Insights Platform / When not to
A quick filter to help you decide if aisot AI Insights Platform is the right fit.
When to use aisot AI Insights Platform
- Portfolio managers at active funds who want to systematize tactical allocation without giving up their own investment style
- Family offices and independent or external asset managers (EAMs and IAMs) that have no in-house quant team but want professional quant infrastructure
- Corporate treasuries focused on protecting asset value against volatility and deep drawdowns
- Quantitative researchers and data scientists at larger asset managers looking for additional signals to feed into existing models
- Structurers who want to turn a validated strategy into a tradable Swiss structured note without building the issuance chain themselves
When not to use aisot AI Insights Platform
- Retail and individual investors: the platform is addressed exclusively to institutional and professional market participants
- Anyone who wants to evaluate a tool self-service, since there is no public price, no online sign-up, no free plan and no announced free trial
- Teams that need published security documentation before procurement, as no trust centre, certification list or hosting location is available
- Buyers with strict GDPR paperwork requirements, because no DPA, retention period or Article 27 representative is published
- Users looking for a mobile or offline product: the platform is browser-based only, with no iOS or Android application
How to use aisot AI Insights Platform
A typical end-to-end flow, from setup to results.
- Start from the product page, where a short video walks through the platform
- Book a call or request a demo: this is the only entry point, as there is no self-service sign-up
- Discuss scope with the sales team and agree a subscription, since access is quoted rather than listed
- Open the AI Insights Dashboard in a browser; nothing is installed locally
- Define your asset universe and build a first portfolio with the portfolio builder
- Set your constraints, including investment restrictions and ESG components, then run the optimizer
- Backtest and iterate with unlimited runs, using the statistical suite to compare volatility and risk-adjusted returns
- Read the agentic AI commentary to understand which factors drive performance and risk
- Move into production: monitor the portfolio and rebalance as the signals update
- Optionally spend the bundled structuring credit to unlock extra features, or issue the strategy as a structured note via the Product Launch Pad
Pros & Cons
Pros
- Covers the whole lifecycle in one place, from strategy and backtesting through to rebalancing and monitoring
- Unlimited runs mean exploration is not metered, which changes how freely a team can test ideas
- Time-boxing the language models tackles look-ahead bias, a real and rarely addressed flaw in backtested research
- The roadmap is genuinely influenceable: clients vote on upcoming features and can commission bespoke ones
- The issuance chain is unusual for an analytics tool, taking a strategy all the way to a tradable security
- Strong academic footing, with an identified ETH Zurich lineage and a named team and advisory board
- A Swiss publisher whose legal name and registered address are published and independently verifiable
Cons
- No public pricing at all: no rate card, no range, not even an order of magnitude
- No self-service path, no free plan and no announced free trial, so every evaluation starts with a sales call
- An API is sold as an available feature yet no documentation page exists anywhere on the site
- No security or trust page: no certification, no named hosting country, no subprocessor list
- The privacy policy is a generic template that covers only the marketing website, never the platform or the portfolio data placed on it
- The site contradicts itself on the company's founding year, giving 2019 on the company page and 2021 in an August 2026 press release
- No contact email address is published anywhere, so all pre-sales exchanges go through a form or a booking link
Pricing & Plans
No price is published. The site has no pricing page at all, and a full review of its sitemap confirms none exists. The vendor does describe a subscription model, stating that subscribing to the platform grants access to its features and that a structuring and feature-request credit is included, but no amount, billing period or tier is disclosed. Access is quoted after a commercial conversation: every call to action on the site leads to booking a call or requesting a demo. Neither a free plan nor a free trial is announced. One figure does appear on the site, USD 10,000, but it is the minimum investment in an actively managed certificate distributed through banks and brokers, and it is not a price for the software.
Data, GDPR & hosting
A consolidated view of how aisot AI Insights Platform handles your data.
GDPR overview
There is no mention of the GDPR anywhere on the site. The privacy policy never names the regulation, and states no legal basis for processing, no data subject rights (access, rectification, erasure, portability, objection), no retention period, no Data Processing Agreement, no named subprocessors, no data protection officer and no representative under Article 27. It also discloses its own origin, having been written from a generic template, and it covers only the marketing website rather than the platform itself. The one age statement, a minimum of 13 years, follows the American COPPA standard rather than European practice. The publisher is established in Switzerland, outside the European Union, and its terms of use are governed by Swiss law. Anything GDPR-related must be obtained directly from the vendor.
Who owns the data?
The published documents are unusually narrow: the privacy policy explicitly covers only the aisot.com marketing website, not the platform, and the terms of use govern the site rather than a customer account. Consequently no clause states who owns portfolio data uploaded to the platform, nor what aisot may do with it. The terms do assert aisot's ownership of the site's intellectual property, and a dedicated Ideas clause states that any feedback or improvement suggestion becomes aisot's property once incorporated. The only product-side statement is the homepage claim that data and assets remain under the customer's complete control. Ownership of client data should therefore be settled contractually before any deployment.
Reuse rights
Nothing on the site grants or restricts a user's right to reuse what the platform produces, so no permission regime can be described from published sources. What the privacy policy does cover is the website: it collects name, telephone number and postal address through forms, plus log data such as IP address, browser version, pages visited and time spent, and states these are used to contact or identify the visitor and to improve the site experience. It commits not to share information except as described, and mentions unnamed third-party companies acting on aisot's behalf. Essential and performance cookies are used; functionality and advertising cookies are declared unused. Reuse rights over signals, backtests and optimized portfolios are a contractual question the site leaves open.
Data retention & training
Hosting summary
The site says almost nothing about hosting. No country, no region and no cloud provider is named anywhere, and there is no security or trust page that would carry that information. The only product-side statement is that customers operate in a secure cloud environment, with data and assets remaining under their control. The AI made in Switzerland line in the footer is a claim about where the technology originates, not about where data is stored, and should not be read as a hosting guarantee. What can be established independently concerns the marketing site rather than the platform: it is published through HubSpot and served behind an anycast content delivery network, which says nothing about where customer portfolio data resides. The publisher is established in Zurich and its terms of use are governed by Swiss law, which sets the contractual jurisdiction but not the hosting location. Any buyer with data residency requirements will have to obtain this in writing.
Things to keep in mind
Risks and trade-offs to weigh before adopting aisot AI Insights Platform.
- The headline gains, up to 90% of time saved, drawdowns down by up to 22% and a higher Sharpe ratio, are claims made by the publisher itself, with no methodology, test period or investment universe disclosed; treat them as marketing until proven on your own data
- Delegating allocation decisions to a signal engine erodes judgement over time; the vendor's own framing keeps the human deciding, and that discipline is worth preserving rather than quietly abandoning under time pressure
- The site contradicts itself on the company's founding year, giving 2019 on one page and 2021 in a 2026 press release, which is a reminder to verify factual claims rather than assume the site is internally consistent
- Nothing is published about whether customer data is used to train models, so this must be settled in writing before any portfolio data is uploaded
- No hosting country is stated anywhere; the AI made in Switzerland claim describes where the technology originates, not where data is stored
- One page of the site, the financial products FAQ, is years out of date, referring to a subscription window that closed in August 2023, and describes certificates rather than the software
- The USD 10,000 figure on that page is a minimum investment in a financial product bought through a bank, not the cost of the software, and the two are easy to confuse
Setup & Integrations
Technical difficulty
There is nothing to install: the platform runs in a browser. The real difficulty is not technical but professional, since using it well requires fluency in portfolio construction, constraint setting, backtesting and risk metrics, although the agentic layer does report its analysis in plain language. Onboarding is not described publicly, and access only begins after a commercial process. Two more technical routes exist, downloading the language models into your own machine learning workflows and using the backtesting and rebalancing API, but neither is documented publicly, so their effort cannot be estimated from the site.
Deployment
Behind aisot AI Insights Platform
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
Who is the aisot AI Insights Platform designed for?
Which markets and forecast horizons are covered?
What does the platform actually contain?
How much does it cost?
Is there a free trial or a free plan?
Is there an API?
What does time-boxing mean in aisot's language models?
Can a strategy be turned into an investable product?
How do I contact the publisher?
Should you pick aisot AI Insights Platform?
The aisot AI Insights Platform is a narrow, serious product rather than a general-purpose assistant. Its value proposition holds together unusually well: quantitative signals, constrained optimization and backtesting in one workspace, and then, rarely for an analytics tool, an issuance route that turns a validated strategy into a tradable Swiss security. The time-boxing of its language models is a substantive technical argument, not marketing varnish, because look-ahead bias is a genuine and under-treated flaw in backtested research. The publisher's ETH Zurich lineage, its named team and advisory board, and three publicly documented funding rounds all support the impression of a company that knows its field. Against that stands almost complete commercial opacity. There is no price, no tier, no trial, no self-service entry point, and no email address. Public documentation is thinner still: the API is sold as a feature yet documented nowhere, there is no security or trust page, no hosting country, no subprocessor list, and the privacy policy is a generic template that covers the marketing site rather than the platform or the portfolio data entrusted to it. The performance figures on the homepage, up to 90% of time saved and drawdowns cut by up to 22%, are the vendor's own claims, published without methodology or test period, and cannot be verified from the site. The practical consequence is straightforward. This is a tool a professional investment team can only assess through a demonstration and a contract negotiation, where pricing, data ownership, model training practice, hosting jurisdiction and API scope all have to be settled in writing. For the audiences it names, the underlying capability looks worth that conversation. For anyone who needs to evaluate before talking to a salesperson, the site simply does not provide enough.
- Choosing a selection results in a full page refresh.
- Opens in a new window.