We’re back with Startup Radar, spotlighting another exciting cohort of Seattle-area founders at the helm of up-and-coming early stage companies.
This batch spans smart home hardware, dementia care, hiring, HR, and AI assistants.
We’re pleased to welcome Aaron Fleishman, partner at Seattle-area venture capital firm Tola Capital, who provided the “VC View” on each startup.
We also decided to mix it up with the addition of “Mean VC,” a custom GPT built by Yohei Nakajima, a Seattle-area investor we wrote about this week. It’s designed to help provide tough feedback for founders on their pitches — so we asked it to judge this week’s Startup Radar companies based on the descriptions below.
Thanks for reading. Email me at [email protected] with potential startups to feature in future Startup Radar spotlights.
Aura Labs
Founded: 2024
The business: Smart home startup initially focused on lighting. Its technology uses sensors and AI to learn user habits and create personalized configurations without the need for technical know-how. The company’s first product is Aura TV Backlight, which aims to create a more immersive home entertainment experience.
Leadership: CEO Josias Garcia was an early sales member at Wyze, the Seattle-area smart home device giant. His co-founders Abu Hafiz (worked at Amazon on smart home products) and Darak Chen (hardware design expert) both stepped down in July.
VC View from Aaron Fleishman: “The Aura TV Backlight appears to be an interesting initial wedge into the home entertainment system. I can definitely see how it creates a more immersive experience for the consumer. The real question I have is: with that initial wedge, where does Aura Labs go? Likening to the B2B software world, what is their land and expand? A whole line of products that make your entertainment experience much more immersive could be really compelling, including technologies like VR for example.”
Mean VC: “Your product taps into the booming smart home market with a sleek, low-friction user experience that could win over tech-averse consumers. Unfortunately, ‘personalized lighting’ sounds more like a Kickstarter gimmick than a defensible business, and you’re one firmware update away from irrelevance when Philips Hue sneezes.”
MeeBoss
Founded: 2025
The business: Job-matching platform aiming to help make hiring faster and more personal for both employers and candidates. MeeBoss offers direct messaging and a job-matching recommendation system with a keyword-based approach. It is currently offering its product to VC-backed startups for free.
Leadership: Co-founder Susanna Niu previously started a news media platform for overseas students and has marketing experience. Co-founder Forrest Fisher was a senior product manager at FloSports and spent more than a decade in tech roles at Disney.
VC View from Aaron Fleishman: “Having invested in talent sourcing automation startups like Fetcher, I see real value in companies that can accelerate the time for the right candidates to be matched to the right companies. Where I think this gets really interesting is when you can turn the company side almost entirely agentic. That could create substantial value on both sides of this talent matching equation.”
Mean VC: “Offering VC-backed startups a free, streamlined hiring tool with direct messaging is a clever wedge into a noisy market desperate for better talent pipelines. But a keyword-based recommendation system in 2025 is like showing up to a gunfight with a butter knife — LinkedIn, AI-native ATSs, and vertical platforms will eat you alive unless you evolve fast.”
Nexhelm
Founded: 2025
The business: AI agent orchestration layer for financial advisors. The company’s AI assistant listens to client meetings, coordinates tasks, and executes assignments across CRM, email, and other platforms. Nexhelm recently joined the AI2 Incubator in Seattle and is running early pilots with independent RIAs and insurance financial managers.
Leadership: CEO Lona Yu was a machine learning scientist at Amazon working on fraud detection models. CTO Nolan McCafferty was also an engineer at Amazon and more recently was at cloud services startup Omnistrate.
VC View from Aaron Fleishman: “Applying AI technologies and agents to lead prospecting and deal coaching is a no-brainer and we are seeing a lot of these technologies already. We are actually invested in a very successful company in the space called Nooks which is more horizontal in nature. My real question for Nexhelm is if the lead prospecting and deal coaching workflows need to be tailored to specific verticals like financial advisors in this case. I could see proprietary data giving them an advantage over horizontal players, but I would want to see how that data evinces itself in the product or models.”
Mean VC: “Embedding an AI agent into the chaotic workflow of financial advisors to automate grunt work is a smart play, especially with warm leads from the AI2 pedigree. But unless you crack compliance, trust, and tight integrations with legacy sludge like Redtail or Salesforce Financial Services Cloud, you’re just another LLM wrapper shouting into a highly regulated void.”
Remme
Founded: 2025
The business: Software platform focused on reminiscence therapy to help improve mood, boost memory, and reduce medication use for older adults with dementia. Remme’s tools use photos and conversation prompts to help users reminisce regularly and naturally while building connections between family members. The company is gathering feedback at local senior centers.
Leadership: Founder Grace Raper was a lead software engineer at AR training startup Ghost Pacer, which had its assets acquired by Peloton. She later launched SortMyShoebox, a computer vision tool that organizes scanned family photos. Raper was inspired to launch Remme by her grandmother, who spent more than 30 years as a hospice nurse.
VC View from Aaron Fleishman: “Unfortunately, with the aging population, the prevalence of dementia and Alzheimer’s continues to grow. I am no expert on this space, but Remme’s reminiscence therapy sounds like it could improve the quality of life for people suffering with these diseases. I could also see opportunities to leverage multimodal AI to further improve this therapy over time.”
Mean VC: “Tapping into reminiscence therapy with user-friendly tools that strengthen family bonds and improve dementia care is both noble and potentially impactful in a massive aging market. But unless you can prove sustained engagement and measurable clinical outcomes — and not just warm fuzzies — this ends up as another well-meaning app collecting digital dust on an iPad in a nursing home.”
Talvita
Founded: 2025
The business: AI-native talent management platform rethinking traditional performance reviews and ways to unlock employee potential. Talvita tracks and evaluates employee impact in high-ambiguity roles, and provides senior leaders with insights. The startup is bootstrapped and plans to launch a limited access MVP.
Leadership: CEO Caitlin Rollman was a senior director at Seattle sales software startup Highspot and spent nearly a decade at Microsoft. Co-founder Kate Wilson was also at Highspot in a senior director role leading a solution architects team, and was a consultant at Accenture.
VC View from Aaron Fleishman: “Performance review management is definitely a crowded market, with both large incumbent HR platforms offering products, but also certain breakout SaaS startups like Lattice. It’s not entirely clear to me how Talvita aims to differentiate from the competition. I could see there being an AI-native angle here which leapfrogs the rest. Particularly interesting would be how they handle AI agents as employees within companies and their performance.”
Mean VC: “Targeting the broken world of performance reviews with AI that actually understands impact in ambiguous roles is a refreshing, high-upside thesis for modern orgs. But bootstrapping an enterprise SaaS in HR without budget, brand, or buyer trust is a long, lonely climb — especially when every CHRO has already been burned by five ‘insight’ platforms that overpromised and underdelivered.”
Read the full article here
