Tool LaunchVon AIRattle

Rattle Pivots to Von AI: A Multi-Model Intelligence Layer for Revenue Teams

Rattle rebrands as Von AI, a multi-model revenue intelligence layer routing tasks to Claude, GPT, and Gemini. Hit $500K revenue in 8 weeks.

6 min readPublished 2026-04-25

What happened

Rattle Software, the company behind the widely-used Salesforce-Slack integration tool Rattle, has pivoted to launch Von AI — a multi-model revenue intelligence platform that routes different tasks to different large language models based on their strengths. According to VentureBeat, Von uses Claude for high-level reasoning and strategic analysis, GPT for bulk data processing, and Gemini for generating creative assets like decks and reports.

The pivot comes from a position of existing revenue. Rattle had built a mid-seven-figure business as a process automation bridge between Salesforce and Slack. But founder Sahil Aggarwal reportedly saw a larger opportunity: building what he describes as "the next Salesforce" — a foundational intelligence layer that understands the entirety of a company's revenue operations rather than just automating notifications between two platforms.

The early traction is notable. Von reportedly crossed $500,000 in revenue within its first eight weeks of launch, with projections to reach $10 million in its first year. The platform is processing more than 10,000 tasks per week and claims 95% accuracy in predicting deal outcomes.

Why it matters for practitioners

Von's launch represents a structural shift in how revenue intelligence platforms are being built. Rather than training a single proprietary model or building on top of one foundation model, Von treats multiple LLMs as interchangeable components in an orchestration layer — selecting the right model for each task based on performance, cost, and capability. This has implications for how CI and revenue teams evaluate and adopt AI-powered tools.

1. The multi-model architecture is a bet on commoditization. Von's design assumes that no single LLM will maintain a durable advantage across all task types. By routing reasoning tasks to Claude, data processing to GPT, and creative generation to Gemini, Von is building for a world where model selection is an optimization problem rather than a vendor commitment. For CI teams, this signals that the platforms built on single-model foundations may face architectural limitations as the model landscape evolves. The question to ask vendors: are you locked into one model, or can your platform leverage the best model for each use case?

2. The context graph concept changes competitive analysis. Von begins its deployment by building what it calls a "context graph" of a company's entire business — including deal stages, territory definitions, organizational hierarchy, and institutional knowledge. Once constructed, this graph provides the foundation for every analytical task Von performs. For deal intelligence, this means Von can contextualize competitive signals within the specific language and structure of each customer's revenue operation, rather than applying generic models. The 95% deal prediction accuracy claim, if validated, would represent a significant advance over existing forecasting tools.

3. The pivot narrative itself is instructive. Rattle's journey from Salesforce-Slack integration to multi-model revenue intelligence platform illustrates a broader market pattern: companies with existing CRM data access are leveraging that foundation to build AI-powered intelligence layers. This has competitive implications for established players like Gong and Clari, which built their intelligence capabilities on proprietary data moats. Von's approach suggests that CRM integration depth plus multi-model AI orchestration could be a viable alternative to proprietary data collection.

4. Revenue velocity suggests genuine market demand. Reaching $500,000 in revenue within eight weeks is fast by any standard, particularly for a company pivoting from an adjacent category. The trajectory suggests that revenue teams are actively looking for intelligence platforms that go beyond recording and analyzing calls — they want systems that can reason about their entire business context and generate actionable outputs. CI teams that partner closely with revenue operations should monitor whether Von's revenue intelligence approach delivers on its prediction claims, as validated deal prediction would fundamentally change how competitive deals are prioritized and resourced.

Key details

  • Company: Von AI (operated by Rattle Software Inc.)
  • Founder: Sahil Aggarwal
  • Previous product: Rattle — Salesforce-Slack process automation (mid-seven-figure revenue)
  • Multi-model architecture: Claude (reasoning), GPT (data processing), Gemini (creative assets)
  • Revenue traction: $500K in first 8 weeks; projecting $10M in year one
  • Task volume: 10,000+ tasks per week
  • Deal prediction accuracy: 95% claimed
  • Core technology: Context graph — maps a company's deal stages, territories, hierarchy, and institutional knowledge
  • Positioning: "The next Salesforce" — foundational intelligence layer for revenue teams
  • Competitive landscape: Positioned against Gong, Clari, and traditional revenue intelligence platforms

Market implications

Von's emergence adds another entrant to an increasingly crowded revenue intelligence market, but its architectural approach sets it apart. The multi-model orchestration pattern — already established in developer tooling and infrastructure — is now arriving in the revenue technology stack. If Von demonstrates that model routing delivers meaningfully better outcomes than single-model approaches, it could pressure established revenue intelligence vendors to adopt similar architectures.

The competitive dynamics are particularly interesting for CI practitioners. Von occupies an adjacent space to traditional competitive intelligence platforms, but its context graph approach could eventually incorporate competitive data alongside revenue signals. A platform that understands a company's deal stages, win/loss history, and competitive landscape well enough to predict outcomes with 95% accuracy would be a fundamentally different tool than today's CI platforms, which primarily focus on intelligence collection and distribution rather than predictive deal analytics.

For revenue operations leaders evaluating their intelligence stack, Von represents a new category of tool: the AI-native revenue intelligence layer that sits between CRM data and decision-making. Unlike conversation intelligence platforms that derive insights from call recordings, or intent data platforms that surface account-level signals, Von aims to synthesize all available data into a single reasoning layer that can answer strategic questions about the business. Whether the multi-model approach delivers on this ambitious vision remains to be validated, but the early revenue traction suggests the market is ready for this category to exist.

The broader signal for the CI market is clear: the boundary between competitive intelligence, revenue intelligence, and deal intelligence is dissolving. Platforms that can reason across all three domains — using whatever combination of AI models proves most effective — will define the next generation of the intelligence stack.

Related resources