-
About the role
Lead AI Engineer – AGI startup (trading) – Remote up to £250k
Do you have a track record of building distributed systems and streaming data platforms at scale?
Are you excited by the challenge of orchestrating thousands of predictive AI agents in real time?
Do you want to help design the backbone of a system that interacts directly with markets?The challenge
We’re an under 2-year-old VC-backed startup with $30m funding building the most accurate model of the world. Our mission is to create prediction systems that evolve in dynamic environments, ingest information at scale, and allocate capital with maximum efficiency. This includes a trading system.As Engineering Lead, you’ll take ownership of the back-end and data infrastructure that powers this vision. You’ll design and build large-scale ingestion pipelines, hypothesis generation engines, and frameworks for orchestrating predictive AI agents. This is a high-agency role where speed, creativity, and technical excellence are essential.
You will build:
- Large-scale, low-latency data ingestion pipelines and knowledge graph systems
- A hypothesis generation and evaluation engine to drive predictive insights
- Frameworks for orchestrating thousands of predictive AI agents, fault-tolerant and context-aware
- Integration of probabilistic models and market feeds into a unified prediction layer
About you
You’re exceptional at building distributed systems and ML infrastructure, comfortable starting from scratch, and motivated by working at the frontier of what’s possible. You thrive in environments where ideas move fast into production and where your systems directly shape market outcomes.What we are looking for is someone with:
Proven expertise in distributed systems, streaming data, and ML infra (Kafka, Flink, Spark, Ray, Dask, Kubernetes)
A strong understanding of building agent systems, including:
- Implementing the latest techniques (context engineering, single-agent scaling)
- Designing fault-tolerant agents (Temporal, Cadence, Airflow)
- Handling long agent contexts at O(100) steps (LangGraph, CrewAI, AutoGen)
- Working with KV cache for efficient inference (vLLM, FlashAttention, DeepSpeed-Inference)
- Building MCP servers and integrating tool calling (OpenAI MCP, LangChain tool abstractions)
- Experience with LLM streaming (WebSockets, gRPC, SSE)
- Experimentation with voice agents (LiveKit, Pipecat, Daily.co, Twilio Voice)
- Up-to-date knowledge of the AI ecosystem, including frontier lab APIs (non-OpenAI compatible) and reverse-engineering APIs (e.g. OpenAI memory, Sesame, Google Generative Service)
Nice to have:
- Experience with GraphQL (gqlgen, Apollo, Hasura)
- Familiarity with databases (Cassandra, Scylla, Postgres, MySQL)
- Work with workflow engines and messaging systems (Temporal, NATS, RabbitMQ, Kafka) or vector databases in RAG projects (Pinecone, Weaviate, Milvus, FAISS, Vespa)
- Publications in conferences/journals or open source contributions (GitHub, Hugging Face, PyTorch ecosystem)
- History of competing in programming competitions (ICPC, Kaggle, Codeforces, LeetCode contests)
Benefits
This role is fully remote, with the core team based in San Francisco.
There’s a salary of up to £200k plus equity, along with typical startup perks.Please note: You must be eligible to work in the UK, US, or EU to be considered.
Interested?
If this sounds like you, get in touch by clicking the ‘apply now’ button or reaching out directly:- Email: mo@mbnsolutions.com
- Call: 02037736498
