Delo | A Wonder by Deloitte Business

One thing I keep seeing in my work with finance leaders: the biggest untapped asset isn’t AI, it’s their own data. Most finance teams I work with aren't struggling with access to AI, they're struggling with what comes before it and what needs to come after. AI can compress a monthly close from days to hours. It can surface anomalies, run scenarios, generate narratives from raw numbers. The technology is production-ready. But here's what separates organizations actually capturing value from the ones running expensive pilots that go nowhere: - Data quality is the bottleneck, not compute. AI is only as reliable as what feeds it and finance should be leading enterprise data stewardship, curating golden datasets, validating sources, auditing for bias. This is a CFO problem, not a "data team" problem. - Controls are what make AI trustable. Human-in-the-loop isn't a limitation, it's the design pattern that lets you scale. Finance professionals shift from doers to orchestrators, owning the judgment layer no model can replace. - The real unlock is strategic, not operational. When you stop spending 80% of your time reconciling data, you can finally do what finance was always meant to do: tell the story of where the business is going, not just where it's been. If you’re curious about what this could mean for your team or just want to swap ideas about how to move forward, I’d love to connect.

ehh I don't know about this one. Parts of the article I agree with are how spreadsheets are just programs, coding agents can connect to different systems and pull data (which is a gap spreadsheets covered till now), and you can now do these much quicker than exporting data and building formulas. On the flip side I think the article misses the crux of what spreadsheets represent, which is custom business logic for nuanced scenarios. Take accountants for example who rarely building simple MoM variance spreadsheets from scratch. It's usually some kind of complex analysis modelled after their business answering very specific questions for the team. As a simpler example, think about a revenue schedule spreadsheet for a project-based business accounting for milestones, discounts, payment terms, etc. For coding agents to build something like this, it needs an ontology (oh no) or some kind of specific understanding of what each field or term means. Without that you get a generic chart or table that's too surface level (i.e., a useless dashboard) or worse an analysis built on wrong fundamentals or assumptions. This is why there's so much hype around building context layers. Then you take into account how the sanctity of these sheets is maintained today by the finance team, usually a single person ensuring there is consistency and continuity so there's actual value in the data being examined. If you didn't read anything, TL;DR coding agents may lead to fewer spreadsheets as data pipelines become easier, but to replace spreadsheets you need an ontology for finance.

60% of work is lost in context, and AI is lost without it. Digital fatigue reduces employee performance by up to 30%. Missing context leads to over 90% of companies missing out on the true value of AI adoption. 3 hours wasted daily searching and stitching context. Come see how Delo is revolutionizing AI teammates. https://trydelo.com/ Delo | A Wonder by Deloitte Business

We're excited to announce the second panel for AI × Professional Services: From Research to Real-World Deployment on February 26th at Mila. Panel 2: Business Perspectives explores AI in high-stakes professional services, from every side of the table. Nicholas Sosiak, CPA is Chief Financial Officer of Cannara Biotech Inc. (TSXV: LOVE), a vertically integrated producer of affordable premium-grade cannabis and cannabis-derivative products for the Canadian market, where he oversees finances and treasury and serves as a key decision maker across the business. Eleonore Jarry is a General Partner at Brightspark Ventures, one of Canada's leading early-stage VC firms, where she leads investments in transformational companies from creation to Series A. Nessim Boussetta, MBA, CPA is a Deloitte Partner and former finance leader at Hopper who founded Delo | A Wonder by Deloitte Business, a Deloitte-backed AI platform deploying digital teammates inside modern finance teams. Bhavish Beejan is Co-Founder and CEO of Material, the AI-native legal services firm for public companies. He previously practiced corporate law at Osler, Hoskin & Harcourt LLP. 🗓️ February 26, 2026 | 5:00–8:00 PM 📍 Mila - Quebec Artificial Intelligence Institute RSVP: https://luma.com/55xk5fg0 Co-hosted by Material and Brightspark Ventures

Nous sommes ravis d'annoncer le deuxième panel de AI × Services Professionnels : De la Recherche au Déploiement, le 26 février à Mila. Panel 2 : Perspectives d'affaires réunit quatre regards sur l'IA dans les services professionnels complexes. Nicholas Sosiak, CPA est chef de la direction financière de Cannara Biotech Inc. (TSXV : LOVE), un producteur verticalement intégré de cannabis et de produits dérivés du cannabis de qualité supérieure pour le marché canadien, où il supervise les finances et la trésorerie et joue un rôle décisionnel clé à travers l'entreprise. Eleonore Jarry est associée principale chez Brightspark Ventures, l'une des principales firmes de capital de risque en démarrage au Canada, où elle dirige des investissements dans des entreprises transformatrices de leur création jusqu'à la Série A. Nessim Boussetta, MBA, CPA est associé chez Deloitte et ancien leader financier chez Hopper, qui a fondé Delo | A Wonder by Deloitte Business, une plateforme d'IA soutenue par Deloitte déployant des coéquipiers numériques au sein des équipes financières modernes. Bhavish Beejan est co-fondateur et PDG de Material, une firme de nouvelle génération combinant IA et expertise professionnelle, dédiée aux marchés des capitaux. Il a précédemment pratiqué le droit des affaires chez Osler, Hoskin & Harcourt LLP. 🗓️ 26 février 2026 | 17h00–20h00 📍 Mila - Quebec Artificial Intelligence Institute RSVP : https://luma.com/55xk5fg0 Co-organisé par Material et Brightspark Ventures

Join us on Feb 26th at Mila - Quebec Artificial Intelligence Institute to discuss practical AI uses cases and adoption in Finance and Legal.

The right place to start for AI automation in finance is to assume whatever routine process you have will soon be automated and done in real-time, and then figure out how you can get the most out of current AI to do it. Most knowledge work processes were built with human constraints (time / capacity). Using any kind of automation, even AI, inherits the same constraints because the process doesn't reflect another reality. It's why most corporate accounting processes are done in batches. Hence, you see most teams using AI to make the old workflow faster. Code transactions in half the time, process invoices 30% quicker, etc. These are marginal gains on a fundamentally broken system. The Ramp announcement last week said the quiet part out loud that many 'AI for knowledge work' companies are saying: If you could process everything in real-time and in no time, what would you change about how you work? That said, existing tools in finance - both point solutions and ERPs, CRMs, etc. - already have AI and will get better at automation within the confines of the system. Once that hits a wall, high leverage teams will look for an AI layer that ties all the systems together, focused more on SOPs, guidelines, nuances, etc. so their team is a reviewer of the boring stuff. Delo | A Wonder by Deloitte Business