AlphaSense disclosed it had crossed USD 500M in ARR with more than 6,500 customers in an October 2025 post on X, up from the USD 400M ARR and 6,000 customers it announced in March 2025. The company also says it now counts every top global investment bank as a customer, alongside the rollout of an earnings-season workflow product pitched as a "streamlined command center" from prep through post-call analysis.

The headline isn't the ARR number. It's the speed of compounding in a category that, two years ago, was still being dismissed as a feature inside Bloomberg or FactSet.

What the numbers actually say

Self-reported ARR is a Tier 3 source by the INFLXD reader's standards, useful color, not IC-defensible without an audit. But the directional read is clean. ARR moved from USD 400M to USD 500M in roughly seven months while net customer count grew by approximately 500. The implied math, if the disclosures are taken at face value, is that average revenue per customer climbed from roughly USD 67K to roughly USD 77K, a ~15% lift in ACV in under a year.

That's the metric to watch in this category. Logo growth is a marketing number. ACV expansion is the signal that buyers are consolidating spend and that the platform is replacing point tools, not adding to them.

The "domain-specific AI" framing

AlphaSense's October post made the positioning shift explicit: "the race isn't just about AI; it's about trusted, enterprise-ready AI that knows your industry." That sentence is doing a lot of work. It's a pre-emptive defense against three different threats in the same breath: generic copilots from Microsoft and Google, vertical-AI startups attacking from below, and the in-house AI teams every bulge bracket and large hedge fund has stood up since 2023.

The claim INFLXD finds credible: enterprise finance buyers are not going to trust a horizontal LLM with their earnings prep. The claim that needs more evidence: that AlphaSense's moat is durable against, say, a sell-side firm that already has Tegus transcripts via API and a competent ML team. The investment banks named as customers are also the firms most aggressively building internal alternatives. That tension doesn't show up in an ARR number.

The Tegus integration is the real story

The acquisition of Tegus in 2024 folded one of the largest expert-call transcript libraries into AlphaSense's search index. For the buy-side reader, this matters more than the ARR milestone. It means a single query can now traverse SEC filings, sell-side research, and expert-call transcripts in one pass, the workflow that used to require three separate tools and a junior analyst stitching the output together.

The earnings-season workflow announcement in January 2025 is the productized version of that integration. Prep, live call, post-call. That's the loop expert networks live in, and it's the loop transcription speed and accuracy actually matter for.

What we'd want to ask next

• What does net revenue retention look like across the customer base, and how does it differ between the original AlphaSense logos and the inherited Tegus accounts?
• How much of the ARR growth is coming from existing customer expansion vs new logo wins, and what's the split by segment (IB, hedge fund, corporate, consulting)?
• What's the pricing structure for the new earnings-season workflow, and is it bundled or sold as a discrete SKU?
• How is AlphaSense positioning against in-house AI teams at the bulge brackets it counts as customers?
• What's the latency and accuracy profile on live earnings call transcription within the workflow, and who is the underlying provider?

The last question is the one that matters most for the transcription layer of the stack.