Robots losing their limbs 🤖
Jul 24, 2024
Kaley Ubellacker
Market Stirrings 🚩
Here's what the week looked like in pre-seed:
$10.9M
Total Amount Raised
10
Total Funding Rounds
$1.1M
Average Dollars per Round
$3.6M-$10.9M
Estimated Valuation Range
Data aggregated from proprietary research and Crunchbase; valuation estimate based on 10-20% ownership stake.
PRE-SEED/SEED DEALS ARE GROWING
Pre-seed and seed deal amounts are migrating upwards, with a higher volume of deals being made at amounts $1 million or more. Seen in the chart from PitchBook below, the number of deals in the $10-25 million range nearly doubled compared to last quarter.
Good Reads 📖
For the rushed reader …
The Faboratory at Yale University created soft robots that can amputate their limbs, reattach, or fuse together with other bots at will.
Google DeepMind has revealed yet another AI advancement, and this time, it's a novel architecture that could solve the problem of scaling LLMs.
Digital health startups had a tough time in 2023, but the first of half of 2024 is off to a slightly less rocky start.
For the less rushed reader …
SOFT LAUNCH: They’re not making robots like they used to anymore; robots are going soft. The Faboratory at Yale University created soft robots that can amputate their limbs, reattach, or fuse together with other bots at will. One example shows a soft robot whose leg gets trapped under a rock. In response, the robot sends a current through its leg, heating up the corresponding joint, so that it can detach and free itself. The potential applications vary widely across rescue operations, space or underwater exploration, and medical use. The innovation boils down to the robot's joints: they were created using a combination of bicontinuous thermoplastic foam and a sticky polymer, allowing the joint to be heavily manipulated, whether that be melted, separated, or reconnected. Yale is so good at research that it's making robots weak in the knees.
MORE AI AP(PEER)S: Large language models just got larger. Google DeepMind has revealed yet another AI advancement, and this time, it's a novel architecture that could solve the problem of scaling LLMs. Mixture-of-Experts (MoE) is a popular technique that scales LLMs without hefty computational costs. However, MoE is limited to a small number of experts, subsets of the model that get routed parts of the data, and it thereby faces computational and memory bottlenecks. DeepMind's shiny new Parameter Efficient Expert Retrieval (PEER) would enable millions of experts to be used, improving the efficiency and scalability of LLMs. You can read the nitty gritty details in the research paper here. Isn't it a little ironic that DeepMind has ~millions of human experts that are pumping out millions of LLM experts? It's AI inception… an AI-ception.
RECOVERY OR DEAD BATTERY: All the rage of pandemic times, digital health startups are struggling like a C-list celebrity to fight their way back into the spotlight. Digital health startups had a tough time in 2023, but the first of half of 2024 is off to a slightly less rocky start. Digital health startups closed $5.7 billion so far in 2024, which is slightly less than H1 2023 at $6.1 billion. However, the number of deals shows promise, ticking up to 266 closings for H1 2024 compared to 244 last year. If the pace maintains, there's potential that digital health startups could raise more money than last year. The good news is that 84% of the deals closed so far in 2024 were early stage, which is seen as a key indicator that venture may be returning to "normal," i.e., sustainable patterns. The digital age isn't getting old, so perhaps it's just a matter of time.
Fire Up the Pre-Seeds🔥
Highlights from this week’s pre-seed raises:
Edtech
Bridgecare - A new homework assignment to take care of your health.
Bridgecare raised an undisclosed amount from unknown investors on June 30. The company is developing an AI-driven platform to help improve students’ mental health, acting as an on-demand support system that will be delivered by certified (human) coaches.
Healthtech
Cell BioEngines - You can stand under my umbilicus.
Cell BioEngines raised $250k at a $13 million pre-money valuation from New York Ventures on July 11. Cell BioEngines is a clinical stage biotech company using umbilical cord blood to create “off-the-shelf” therapies to treat diseases.
Synaptrix Labs - Tech that’s moving quickly.
Raised an undisclosed amount from Underdog Labs and other unnamed investors on July 11. Based in New York, Synaptrix Labs is creating a non-invasive neural interface that allows paralyzed individuals to control their wheelchair using brain activity and eye movements.
Fintech
BucksApp - Can you bank on your data?
BucksApp raised $1 million led by iThinkVC and BuenTrip Ventures on July 9. BucksApp targets LatAm banks and is developing data processing technology to optimize transactional data and enhance sales strategies.
SKROL - For business owners hoping to chill on an island in the Caribbean instead.
SKROL raised $1.2 million led by Magnetic Capital on July 8. For retiring business owners, SKROL is using AI to acquire and transform SMBs, aiming to acquire 5,000 businesses over the next 4 years. It was founded by Michael Cassau, who is onto his second venture after founding Grover, a marketplace for renting technology.
E-commerce
Garnet Gazelle - Specializing in the family jewels.
Garnet Gazelle raised $100k from Keyhorse Capital on July 9. Based in Kentucky, Garnet Gazelle is building an auction site specialized in jewelry and gemstone liquidation, using pre-bidding to keep the ball rolling.
RFPrime - The RFP process used to be clunky; now, it’s like floating on a cloud.
RFPrime raised an undisclosed amount from unknown investors at an $880k pre-money valuation on July 12 and is based in Houston. Instead of passing around multiple documents, work on an optimized, single cloud-based version.
Artificial Intelligence
Thunder Compute - The future of computing just got cloudier.
Thunder Compute raised $500k from Y Combinator at a $7.1 million pre-money valuation on June 9. Based in Atlanta, Thunder Compute wants to solve the GPU shortage by making it easier than ever to access local development environments.
Niva - Sending agents to the border.
Niva raised $3.3 million led by Gradient Ventures, Google’s AI-focused venture fund, on July 8. Niva is transforming global commerce with compliance and fraud AI agents that streamline international requirements for cross-border transactions. The tech reduces onboarding cycles for fintech companies from weeks to months.
Outro🚪
Have any questions, feedback, or comments? Just reply here. We iterate and curate the newsletter according to your interests!
Some last matters of business:
If you’re a technologist (engineer or product manager / designer with a technical background) join us on the NVTC LinkedIn group if you haven’t. We’d love to have you!
Sign up here if you’re interested in co-investing with Necessary.
If you’re a startup founder, we’d love to help where we can! Brex provides full-stack finance solutions for startups. Sign up via Necessary and get bonus points.
Thanks for reading, and see you next week!
