Aside Gives Sales Reps the Answers They Need While the Prospect Is Still Talking

The Macro: Sales Calls Are Where Knowledge Goes to Die

I sat in on a demo call last month where a sales rep got asked a straightforward question about API rate limits. He did not know the answer. He said “let me get back to you on that” and moved on. The prospect’s energy dropped immediately. The deal closed eventually, but it took two extra weeks and a follow-up call that should not have been necessary.

This happens constantly in enterprise tech sales. The product is complex. The documentation is scattered across Notion, Confluence, Slack threads, and the brains of engineers who do not join sales calls. Reps are expected to know everything about a product that changes every sprint. They do not. They cannot. And every time they say “let me circle back,” the close rate drops.

The existing solutions are not great. Gong records calls and gives you post-call insights, which is useful but does not help you in the moment. Chorus does something similar. Clari focuses on forecasting. The real-time coaching space is newer and less crowded. Cogito does real-time emotion analysis. Observe.AI does quality assurance. But none of them are solving the specific problem of “the prospect just asked a technical question and the rep does not know the answer.”

This is a retrieval problem, not a coaching problem. The answer exists somewhere in the company’s knowledge base. The rep just cannot find it fast enough while simultaneously maintaining a conversation. That is the gap.

The Micro: Founding Engineers From Airbridge Built Something They Needed Themselves

Aside listens to sales calls in real time and surfaces answers from internal knowledge sources. Documentation, Slack conversations, HubSpot data, and recordings of past calls where someone already answered the same question. The system pulls relevant information and presents it to the rep while the prospect is still talking.

Jun Kim is the CEO. He was a founding engineer at Airbridge.io, which scaled to $30M ARR. Before that, he worked as a solutions consultant, which means he was literally the person answering technical questions on sales calls. He built Aside because he lived the problem. Chanhee Lee, co-founder, was also a founding engineer at Airbridge. Sanghun Lee, the third co-founder, previously built Caret, where he developed real-time audio pipelines for AI meeting notes. The audio processing expertise is directly relevant here. Getting low-latency transcription and retrieval working reliably during a live call is a hard engineering problem, and having someone on the team who has already built real-time audio infrastructure is a meaningful advantage.

They are a three-person team out of San Francisco, Y Combinator Fall 2025 batch.

The feature set goes beyond just answer retrieval. Aside includes pre-configured “Live Cards” with instant access to FAQs and playbooks, live coaching based on frameworks like BANT and SPICED, and post-call feedback for continuous improvement. The coaching layer is interesting because it means the product is not just a search engine for your docs. It is watching the conversation flow and flagging when the rep misses an opportunity to qualify budget or establish timeline.

The competitive positioning is smart. Gong owns the post-call analytics market. Aside is not trying to compete there. Instead, they are going after the real-time moment, the 30 seconds between when a question is asked and when the rep needs to respond. That is a narrow wedge, but it is a wedge that directly impacts revenue. If Aside can demonstrably reduce the number of “let me get back to you” moments on sales calls, the ROI math writes itself.

The technical challenge is latency. The system needs to transcribe speech, understand the question, search multiple knowledge sources, and present a relevant answer, all within a few seconds. If it takes 15 seconds, the moment has passed and the rep has already improvised an answer or punted. The team’s background in real-time audio systems suggests they understand this constraint, but scaling it across thousands of concurrent calls with acceptable latency is a different problem than making it work in a demo.

The Verdict

I like this product because the problem is concrete and the solution is measurable. Either reps answer more questions on calls or they do not. Either close rates improve or they do not. There is no hand-waving about “alignment” or “synergy” required to understand the value proposition.

At 30 days, I want to see average response latency in production. Under three seconds is the threshold where this is genuinely useful. Over five seconds and it is a novelty. At 60 days, I want to know how accuracy holds up across different knowledge bases. A company with clean documentation in Notion will get better results than one with critical information buried in Slack threads from 2024. At 90 days, the question is whether sales leaders are buying this for entire teams or whether it stalls at individual rep adoption. The difference between a tool one rep loves and a tool the VP of Sales mandates is the difference between a small business and a real company.