This week, we explore our lead investment in Spara’s $15 million seed financing round. Co-founded by CEO David Walker and CTO Zander Pease, Spara is an enterprise-grade AI platform that deploys voice, email, and chat AI agents to engage, qualify, and convert leads into revenue instantly.
If you’ve bought software, you’ve been there: you’ve landed on an enterprise platform ready to transact, but get stuck behind a “book a demo” button.
For sellers, this inefficiency hurts. SaaS companies often spend almost half of their revenue on sales and marketing efforts, but are still bottlenecked by human processes: limited sales representative capacity, scheduling constraints, and inefficient workflows that make prospective buyers wait.
Language models offer a new set of solutions. With AI, buyers can experience a faster, more personalized, always-on experience, and sellers can enhance efficiency and accelerate time to revenue.
That’s why we’re thrilled to share that we’ve led Spara’s seed financing to transform go-to-market, alongside Inspired Capital. Spara is an enterprise-grade AI platform for voice, email, and chat AI agents to instantly engage, qualify, and convert leads to revenue. The company gives customers their own dedicated AI assistant that lives on their website and engages buyers the moment they show intent. Spara can qualify leads, answer product questions, and book meetings and demos in real time. Their long-term ambition is to leverage AI to rethink how companies engage with their buyers.
It’s an ambitious mission, and Spara is the team to do it, combining unique domain and company-building expertise. CEO David Walker founded and sold Triplemint, where he led and scaled the company’s sales organization. CTO Zander Pease has been a longtime product and technical leader across companies like Hyperscience and Nomad Health.
Already, they are off to the races. At Rho, the Spara platform booked over 90 meetings in 30 days, converting 50% of qualified traffic with an average chat-to-meeting-booked time under three minutes. And they’re just getting started in augmenting sales teams with AI, working on expanded capabilities, more integrations, and further enterprise controls.
We’re thrilled to partner with David, Zander, and the Spara team to transform buying and selling with AI.
Read the company’s feature on Business Insider to learn more.
AI News This Week
-
Radical Ventures Closes $650-million Early-Stage AI Fund (The Globe and Mail)
Radical Ventures closed its latest early-stage fund. The over-subscribed US$650 million fund marks Radical’s fourth early-stage fund and sixth fund overall, following the US$800 million growth fund that was launched last year. Radical has a track record of backing many top AI startups at the early stage, and looks forward to continuing that success via Fund 4.
-
The RL Renaissance (Radical Talks)
This week on Radical Talks, Radical Partners Molly Welch and Aaron Rosenberg discuss the resurgence of reinforcement learning in AI. They speak with Ross Taylor, who led Meta AI’s reasoning team during the development of Llama 2 and 3 and co-founded General Reasoning. Ross makes several provocative points, suggesting that current software engineering benchmarks for AI are “too artificial,” as studies have shown that few AI-generated pull requests would be accepted in real-world workflows. He also states that the obsession with AI for math and code misses economically valuable domains. Ross emphasizes that despite RL’s growing power, humans remain essential for providing domain expertise and ensuring reliability.
-
The Best Inventions of 2025: FireSat (Time)
Radical Ventures portfolio company Muon Space, in collaboration with Earth Fire Alliance and Google Research, is deploying FireSat, a satellite constellation that detects fires as small as five square meters. Named one of TIME’s Best Inventions of 2025, FireSat scans fire-prone regions every 20 minutes using advanced infrared sensors. Nine fire agencies across three continents are now using the platform. Muon Space was also awarded a $44.6 million U.S. Space Force contract for dual-use environmental monitoring technology based on FireSat’s capabilities.
-
AI Video: More Than Just “Slop” (The Economist)
Video AI models are emerging as the next frontier in artificial intelligence. Beyond the much-hyped content creation, the true potential of video models lies in their problem-solving abilities. Researchers have demonstrated that advanced video models can perform visual and spatial tasks without specific training, including edge detection, image deblurring, solving Sudoku puzzles, and planning robotic motions. These systems work by progressively “de-noising” visual static, asking at each step what would make the output match the given prompt. Under this paradigm, video models act as “zero-shot reasoners” capable of handling diverse visual challenges through “chain-of-frames visual reasoning.”
-
Research: Document Summarization with Conformal Importance Guarantees (Signal1)
Researchers from Radical Ventures portfolio company Signal 1 have developed Conformal Importance Summarization, a breakthrough framework that provides statistical guarantees for AI-generated document summaries to include critical content. Existing systems lack reliable assurances that important information is preserved, a crucial limitation in high-stakes fields like healthcare, law, and finance. This new method uses conformal prediction to offer rigorous, distribution-free coverage guarantees by calibrating thresholds on sentence-level importance scores. The approach enables extractive document summarization with user-specified coverage and recall rates over critical content, ensuring that summaries retain essential information with mathematically proven reliability.
Radical Reads is edited by Ebin Tomy (Analyst, Radical Ventures)
