Backend Developer

About the Project

TODS Ventures is developing a highly ambitious Enterprise AI Governance Platform — a research-led project at the intersection of artificial intelligence, mathematics, law, ethics, and enterprise technology. The platform sits as a managed gateway between corporate users and external AI providers, applying rigorous, multi-dimensional governance to every AI interaction at scale. The project combines original academic research with practical software engineering, and is preparing for peer-reviewed publication alongside a commercial product launch. This is an early‑stage, high‑calibre founding team engagement.

The Role

You will be the sole Backend Developer on the founding team — the engineer who builds the entire server‑side stack from the ground up. The platform is technically sophisticated: it sits as a high‑performance AI gateway, applies complex multi-dimensional governance logic in real time, maintains an immutable audit trail, manages knowledge retrieval from a vector‑based repository, and supports async human review workflows.

The stack spans Python, Node.js, Java, and Rust , with deep integration with Anthropic and OpenAI APIs throughout. You will work directly with the CEO/CTO and collaborate with the Frontend Developer on API contracts. We are open to both full‑time and part‑time arrangements for the right candidate.

Key Responsibilities

• Design and build a high‑performance LLM gateway and proxy — intercepting, routing, and governing AI requests across multiple providers (Anthropic, OpenAI, Azure, Google) with minimal latency overhead.
• Implement a governance and decision engine — applying complex multi‑dimensional rules and policies to AI interactions, producing structured outcomes with full audit logging.
• Build a security and data protection layer — data loss prevention, content filtering, prompt injection defence, and cryptographic non‑repudiation.
• Develop a vector‑based knowledge retrieval system — embedding, indexing, and querying structured governance content using modern vector database technologies.
• Build a real‑time observability and analytics pipeline — statistical monitoring, drift detection, and composite scoring across large volumes of AI interactions.
• Implement an async multi‑party workflow engine — managing human review queues, structured deliberation, quorum logic, and tamper‑evident case records.
• Build enterprise connectivity integrations via MCP — connecting the platform to CRM, ERP, HR, and other enterprise systems.
• Own AWS cloud infrastructure — EC2, RDS/Aurora, S3, API Gateway, CloudFront, CloudWatch, IAM.

Required Skills & Experience

• Python (FastAPI or equivalent) — 3+ years building production APIs and data pipelines.
• Vector databases — pgvector, Pinecone, Weaviate, or equivalent; embedding pipelines.
• PostgreSQL — advanced query design, indexing, migrations, audit logging patterns.
• Authentication & security — SAML 2.0, OIDC / OAuth 2.0, JWT, mTLS.
• Docker, CI/CD (GitHub Actions), infrastructure‑as‑code basics.
• Strong understanding of API gateway patterns and high‑throughput proxy design.

Desirable / Nice to Have

• Java — enterprise integration patterns, Spring Boot or Micronaut.
• Rust — systems‑level performance work, cryptographic operations.
• Knowledge graph design and graph embedding models.
• MCP (Model Context Protocol) — building or consuming MCP servers.
• Statistical methods, linear algebra, or applied mathematics background.
• Comfort reading and translating academic research papers into implementation.
• DLP, data redaction, or content classification system experience.
• Familiarity with GDPR, EU AI Act, or financial services regulatory requirements.

Duration 1 Jul 2026 – 31 Mar 2027

Commitment Full-time or Part-time

Location UK or EU · Remote / Hybrid

Rate To be negotiated

IR35 Note: This engagement is expected to fall outside IR35. An independent IR35 status assessment will be required before work commences. Applicants should operate via their own limited company.