onepot provides an AI‑driven retrosynthetic planning platform that generates ranked synthesis routes directly from user‑submitted molecular structures. The cloud service evaluates reaction feasibility, cost, and step count while keeping all data encrypted end‑to‑end, and includes a web‑based structure editor and RESTful API for LIMS/ELN integration. It serves pharmaceutical R&D teams, biotech firms, CROs, and academic labs that require confidential, high‑throughput synthesis design.
Funding
Funding not disclosed
Founders
Product
Problem
Designing synthetic routes for small‑molecule drug candidates requires extensive cheminformatics expertise and computational resources, yet many organizations lack secure, in‑house tools. Confidentiality concerns arise because proprietary molecular structures must often be shared with external software providers, exposing intellectual property to potential leakage.
Solution
onepot delivers an AI‑driven retrosynthetic planning platform that generates viable synthesis pathways directly from user‑provided structures. The service runs advanced machine‑learning models on cloud infrastructure to evaluate reaction feasibility, cost, and step count, returning ranked routes in minutes. All molecular data are encrypted in the browser using RSA‑OAEP and AES‑256‑GCM, transmitted over TLS, and protected by AWS KMS, ensuring that the provider never accesses plaintext structures. Users interact through a web‑based structure editor that supports similarity and substructure queries, enabling rapid iteration without leaving the secure environment. The platform’s API can be integrated into existing LIMS or ELN systems, allowing seamless workflow automation while maintaining end‑to‑end data confidentiality.
Target Audience
The primary customers are pharmaceutical R&D teams, biotech firms, and contract research organizations that require confidential, high‑throughput synthesis planning. Academic chemistry groups seeking secure access to state‑of‑the‑art AI retrosynthesis also benefit from the platform.
Features
- Browser‑side encryption of input structures with RSA‑OAEP key exchange and AES‑256‑GCM payload protection
- Cloud‑hosted deep‑learning retrosynthesis engine that scores routes by yield, step economy, and reagent availability
- Real‑time similarity and substructure search to suggest alternative scaffolds or precursors
- Zero‑knowledge architecture: encrypted data never appear in plaintext on the server, complying with IP protection policies
- TLS‑secured communication and AWS Key Management Service for automated key rotation and audit logging
- Web‑based structure editor with drag‑and‑drop editing, atom‑level annotation, and export to common chemical file formats (SMILES, MOL)
- RESTful API and SDKs for integration with laboratory information management systems (LIMS) and electronic lab notebooks (ELN)