Innospark Ventures
31 Saint James Ave Ste 740, Boston, Massachusetts, 02116, United States
Overview
Innospark Ventures is a new breed of early-stage venture capital firm, investing in the AI-driven economy.
- Total investments
- 26
- Lead investments
- 10
- Investments · 12mo
- 1
- Active investors
- 4
Sector focus
- Artificial Intelligence (AI)
- B2B
- Cyber Security
- Education
- Information Technology
- Life Science
- Machine Learning
Investment portfolio
- Fourier Health
Participated · Seed · Oct 2025
Founded in late 2023 by James Lloyd and Christopher Lee, Fourier Health offers an AI-driven platform that ingests PDFs, faxes, handwritten notes and other heterogeneous data sources, integrates with multiple EHR systems, and produces concise, context-specific clinical summaries. Its clinician-in-the-loop architecture pairs large-language-model outputs with specialist review to ensure accuracy and continuous model improvement. Providers using the product report over 98% reduction in document validation and the recovery of 2–3 hours of administrative time per patient, improving onboarding and care coordination. The platform also automates document labeling and harmonizes data to give a longitudinal view of patient wellness. With fresh capital, the company plans to expand its engineering, sales and implementation teams and continue refining its R&D infrastructure. Fourier’s vision is to embed last-mile, specialty-specific AI into existing clinical workflows, accelerating diagnosis, boosting provider revenue, and enhancing patient outcomes.
- Persist AI
Participated · Series A · May 2025
Persist AI offers an AI-driven robotics platform and Cloud Lab that lets pharmaceutical scientists remotely predict, build, and test formulations using miniaturized, automated laboratory facilities. Its AI models generate formulation recipes and the robotic systems execute high-throughput experiments, dramatically reducing material use and timelines. In a recent engagement, Persist identified an optimal long-acting injectable in two months by building and testing 700 formulations, versus an industry standard of roughly 10–15 formulations per month. The Cloud Lab supports modalities including peptides, small molecules, antibodies, and antisense oligonucleotides, and serves multiple top-10 pharmaceutical clients as well as smaller biotechs. Persist plans to expand its robotic lab, broaden the range of formulations its models can predict and test, and build comprehensive datasets to improve its AI. The company is Y Combinator–backed (graduated 2023) and recently closed a $12 million Series A to fund these efforts. Persist AI is an AI-driven drug formulation company based in Woodland, CA. The company has developed robotics that enable automated development of long-acting injectable formulations. Coupled with AI models, their technology rapidly predicts which formulations will release drugs in the body for an extended period and reduce the number of injections patients need. Led by CEO Karthik Raman, Persist AI completed YCombinator and subsequently raised $4M in seed financing. The round was led by 2048 Ventures with participation from Innospark Ventures, Fellows Fund, YCombinator, Pioneer Fund and others. Persist AI intends to use the funds to expand operations and further its development efforts.
- Copernic Catalysts
Participated · Seed · Nov 2024
Copernic Catalysts uses computational design and high-throughput experimentation to develop transformational catalysts for sustainable e-fuels and bulk chemicals. Its first proprietary catalyst is an improved, drop-in replacement for ammonia synthesis catalysts that can operate at lower temperatures and pressures, reducing CapEx and OpEx. The company is scaling that ammonia catalyst to the kilogram scale and is working with one of the world’s largest ammonia producers to test the technology. Through an exclusive research collaboration with Schrödinger, Copernic plans to develop a second catalyst target for Fischer–Tropsch syngas-to-SAF to improve aviation carbon intensity. Founded in 2021 and based in Cambridge, Mass., the company is backed by venture investors and was the recipient of a $3.58 million ARPA-E OPEN 2021 grant from the U.S. Department of Energy.
- Phaseshift Technologies
Led · Seed · Oct 2024
Phaseshift develops new metallic alloys and composites using its proprietary AI-powered Rapid Alloy Design (RAD)™ platform to generate candidate materials and run multi-scale simulations before physical experiments. The platform aims to shorten development timelines, reduce unsuccessful experiments, and cut upfront investment needed to find viable materials. Phaseshift targets practical, drop-in replacement materials for industries including aerospace, automotive, mining, energy, and advanced manufacturing. The company plans to accelerate development and commercialization of specialized materials that improve durability, efficiency, and sustainability while addressing supply‑chain and raw‑material constraints. Management says the technology enables faster, more accessible materials innovation and helps reduce waste and maintenance costs across sectors. The company is led by co‑founders Fazal Mahmood (CEO), Dr. Gurjot Dhaliwal (CTO), and Dr. Abu Anand (CSO), and is based in Toronto.
- Antiverse
Participated · Seed · Oct 2024
Antiverse combines proprietary generative machine-learning models with programmable cell-line engineering to rapidly create epitope-specific, therapeutic-grade antibodies against difficult targets such as GPCRs and ion channels. Its lab-in-the-loop workflow reportedly reduces de novo antibody development timelines to under four months, integrating in-silico design with in-house build and validation. The company supports partnered discovery programs with several top-20 pharmaceutical firms and has just launched a research collaboration with the Cystic Fibrosis Foundation to generate antibodies targeting the extracellular region of CFTR. Backed by more than $20 million in total funding, Antiverse is expanding its platform, accelerating its internal pipeline, and intends to move its first wholly owned candidates into later-stage preclinical development by 2027. Headquartered in Cardiff, UK, with offices in Boston and Prague, the team has spent seven years training its models on challenging targets to optimize physicochemical properties and humanness of antibodies. Antiverse offers flexible collaboration structures to biotechnology, pharmaceutical, and foundation partners while simultaneously advancing its own therapeutics.