Aily Labs wants the boardroom in your pocket

Consider the journey of a number inside a multinational company. A country team produces it. A regional team checks it. Headquarters reconciles it with other numbers. By the time it becomes a plan, the world that produced it may have changed. The spreadsheet is immaculate. Its timing is less impressive.

Bianca Anghelina knew that routine from inside Novartis, where she led Global Digital Finance. In a Leadership Matters interview, she described planning cycles lasting nine to twelve months. Applying AI brought the process down to weeks. The attraction was practical: people could see a more realistic plan together, while there was still time to use it.

First, automate your own headache

Anghelina started Aily Labs in June 2020. That August, she teamed with Sara Bisbe López, an AI specialist who had also worked at Novartis. Business knowledge and technical knowledge arrived together. The first product took eight weeks, according to their investor’s founder profile; the company then bootstrapped for three years.

The founders wanted an experience with the ease of everyday consumer apps. This is an unusually sensible ambition for corporate software, a category in which inconvenience can pass for seriousness. Aily’s mission is to make AI useful to people across a large organization, including those who have no desire to become data scientists.

A warning is worth more when someone can act

In June 2023, Sanofi announced the company-wide rollout of plai, developed with Aily. The app aggregated internal data across functions and offered personalized what-if scenarios. Sanofi also reported that plai could predict 80% of low-inventory positions in its biopharma supply chain, allowing teams to mitigate shortages.

That is a useful example because the chain of value is visible. An inventory forecast identifies a risk; an employee can respond before supply becomes a problem. The outcome depends on both halves. A warning that arrives without an owner is simply another notification competing for attention.

Aily’s current Sanofi case study reports more than 23,000 global users and $300 million in avoided revenue losses. Those are vendor-reported figures, rather than independently audited results. They nevertheless describe a deployment extending beyond a small experiment. The interesting question becomes how an app earns a place in ordinary work.

“It’s like snackable AI in your pocket”Paul Hudson, quoted in Aily’s Sanofi case study

The decision lives between departments

Aily’s product ladder follows the widening scope of that question. Its App focuses on a single module. Pro connects departments through the company’s Correlator technology. ProX provides personalized agent advisors. The Super Agent coordinates specialized agents and is designed to carry recommendations through to execution.

For a buyer, the distinction matters. A department can improve its own numbers while creating trouble elsewhere. As an illustrative example, reducing inventory might flatter a finance metric while leaving a commercial team unable to meet demand. Connecting the measures makes that trade-off visible before it becomes an argument.

This places Aily among business intelligence, planning software and enterprise AI. Its pitch emphasizes connected decisions, mobile access and action. Conventional reporting tools and internally built agents remain alternatives. The sensible comparison is the complete workflow: which systems must connect, who reviews a recommendation, and who is permitted to execute it?

Take that inventory example a step further. A recommendation to move stock has consequences for transport costs, customer commitments and cash. A useful system would let the decision-maker inspect those consequences together. Otherwise, the organization has merely exchanged a spreadsheet argument for an algorithmic one. This is why the cross-functional promise deserves more scrutiny than the charm of the interface.

The software company hiding inside the services

Early on, Aily did not resemble a tidy venture-capital category. Insight Partners described substantial services revenue and saw potential for recurring software revenue as daily usage grew. Its involvement included help with hiring and building repeatable platforms. The €19 million Series A arrived in August 2023.

What changed Anghelina’s mind about outside funding? In the investor’s account, the accelerating pace of AI innovation made speed more valuable. Independence had funded the early business; capital could help it expand faster. In November 2025, Aily announced an $80 million round led by FPV Ventures, with Insight Partners and J.P. Morgan participating.

These figures describe financing, not the price a customer pays. Aily sells enterprise software with pricing tied to modules, user volume and deployment scope. Integration and enablement belong in that purchasing conversation too. Comparing proposals by seat price alone would miss much of the work required to make a recommendation useful.

For a prospective customer, a useful pilot would follow one decision through its full life: the initial warning, the options considered, the person approving a response and the eventual outcome. That is an evaluation framework, rather than a claim about every Aily deployment. It also helps expose an expensive illusion: activity inside an app can rise while the business remains unchanged.

Getting through the corporate front door

Distribution became more concrete in 2026. Aily announced an AWS partnership in July and a Google Cloud partnership in August, bringing marketplace procurement into the story. Existing cloud commitments can make buying easier. Aily says certain AWS customers can deploy inside their own environment in as little as one day; that is a conditional deployment claim.

Reliability is another part of the proposition. Aily’s December 2024 partnership with Quebec’s Mila research institute targets knowledge graphs and AI systems that assess their own answers and seek more information when uncertain. For an enterprise agent, knowing when to hesitate could be as useful as knowing what to recommend.

There is also a cultural bargain here. Sharing a forecast across functions removes some of the privacy departments enjoy when each maintains its own version of the future. That can make conversations more productive, but only if leaders accept that a recommendation may challenge a local target. For a buyer, testing that willingness alongside the technology would be a useful way to learn whether the proposed workflow can survive the first disagreement.

The practical lesson is available without buying anything: choose a recurring decision, connect the information it needs, give somebody authority to act, and measure the result. This is an editorial inference from Aily’s approach. Disputed metrics, inaccessible data or unclear permissions weaken it. The boardroom may fit in a pocket; responsibility still needs an address.