
1517 Fund
5519 McMillan St, Oakland, CA, 94618, United States
Overview
1517 supports teams with pre-seed and seed funding for technology startups. Their focus is on makers, hackers, and scientists interested in working outside tracked institutions because we believe that the path geared towards higher education is not for all. They motivate people to work on what they’re passionate about, to learn by doing, and to create new technologies.
- Total investments
- 38
- Lead investments
- 5
- Investments · 12mo
- 5
- Active investors
- 4
Sector focus
- Venture Capital
Investment portfolio
- Corner
Participated · Equity · Jul 2026
Founded in 2022, Corner is a social-mapping app aimed primarily at Gen Z users that adds locations to its map only when a user saves them. The platform has over 250,000 users across 425 cities, primarily concentrated in New York and Europe. Corner offers user-generated reviews, photos, curated lists, and features such as bookings/events, reservation alerts, and age- and location-based recommendations. Last year it launched an AI-powered search that uses its proprietary dataset to deliver personalized suggestions. The founders, Eliza Wu and Jake Xia, have helped build the brand presence on social media, and Corner’s digital magazine has more than 91,000 Instagram followers. The company faces competition from larger mapping and review players like Google Maps, Apple Maps, and Yelp while preparing to pursue a Series A.
- Arcturus
Participated · Seed · Jun 2026
Arcturus develops carbon-nanomaterial-infused copper and aluminum conductors engineered as drop-in replacements for traditional metals. Its process uses lasers to infuse carbon nanomaterials into conductors, producing proof-of-concept wire several centimeters long to date. The company says the materials can carry more electricity with less heat loss, potentially cutting grid losses in half and unlocking roughly 3% more electricity on average and up to 10% during peak congestion. Arcturus has been operating in stealth under founder and CEO Amir Mashal, who has been refining the materials in a garage in Malibu, California. With the $8 million seed funding, the company plans to scale production to tens of meters and test the material in motor windings, busbars, drones, robotics, and data centers. The materials are positioned as form-factor compatible with existing copper and aluminum applications to avoid system redesigns.
- Reprogram Biosciences
Participated · Seed · May 2026
Reprogram Biosciences develops mRNA-encoded gene-combination therapeutics intended to reprogram tumor cells in situ to induce antigen-presenting function and activate antitumor immunity. The company uses an AI discovery engine, CellRecodeX, which integrates biological foundation models to identify and prioritize therapeutic reprogramming candidates and support pipeline expansion. Reprogram is focused on treating solid tumors with high unmet need and is operating at the preclinical stage. The team reports accelerated progress—advancing from concept to in vivo data within six months. Current plans funded by the seed round include advancing the lead program, expanding CellRecodeX, and supporting CMC activities. Founded in 2025, the company is based in San Carlos, California and has raised a total of $6 million to date.
- Positron
Participated · Series B · Feb 2026
Founded in 2023 and based in Reno, Nevada, Positron develops energy-efficient AI inference hardware, including its shipped Atlas inference system. Atlas has already been delivered to customers such as Cloudflare, and the company claims its hardware can deliver more than 3.5x performance per watt versus Nvidia's latest products on certain inference workloads. Positron is developing a next-generation chip called Asimov, targeting tape-out in late 2026 and production in early 2027. The company emphasizes U.S. manufacturing and energy efficiency as core differentiators. Positron has raised roughly $309 million to date, including a $230 million Series B in February 2026 that valued the company above $1 billion. CEO Mitesh Agrawal has led the company since before its seed round.
- Astrus
Participated · Seed · Sep 2025
Astrus develops AI software to automate the manual and time-consuming process of analog circuit layout in semiconductor design. The company applies reinforcement learning to generate thousands of circuit layouts from schematics and performance requirements. Founded in 2023 by Brad Moon and Zeyi Wang, Astrus is headquartered in Toronto and Kitchener–Waterloo, Canada. It raised $8.0M USD (CA$11.0M) in seed funding to expand its team and advance R&D. The company plans to use the proceeds to prepare for paid pilots with Tier-1 semiconductor companies in 2026. The founders have stated an ambition to build something lasting in Canada rather than relocate or pursue a quick exit. Astrus builds AI software that automates analog chip layout design, allowing circuit designers to input schematics and receive manufacturable physical layouts almost instantly. The product automates the traditionally manual layout step—currently a months‑long iteration loop—reducing iteration time to as little as a day and letting designers focus on higher‑level work. The company does not plan to tape out chips itself; instead it aims to democratize analog design and enable creation of analog circuits for digital chips, analog AI chips, and neuromorphic applications. Astrus emphasizes training machine‑learning models to capture design intuition and manufacturing constraints so generated layouts are manufacturable. The founders position the market as high value—about 60,000 analog designers globally and a multi‑billion dollar revenue opportunity. Astrus was founded in 2022 by Brad Moon and Zeyi Wang and has disclosed early-stage fundraising activity. Astrus develops an AI engine to automate analog IC placement and routing, aiming to dramatically shorten manual layout cycles and boost designer productivity. The company has built a deep reinforcement learning prototype (Astrus AI) and is demonstrating the PoC to potential customers with positive feedback. Astrus plans to commercialize its first product in about four to eight months after finishing its technical PoC and is optimizing current offerings for certain analog chip types and technology nodes. Longer-term the company intends to generalize its AI engine to support all circuits and technology nodes and to launch a specialized product class for the high‑speed market. Astrus estimates a large market opportunity and the founder projected a $2 billion annual revenue scenario based on industry designer counts and pricing. Development work has focused on rebuilding an AI-friendly layout environment and addressing parasitic effects, which the company cites as a competitive advantage.