What are the responsibilities and job description for the Applied AI Engineer position at Base0?
About Base0
We’re solving one of the biggest challenges in modern AI workflows: fragmented context. Today, project knowledge is scattered across conversations, tools, and docs—forcing teams to spend more time steering AI than actually getting work done.
Base0 brings context, continuity, and observability to AI-powered work, enabling teams to ship faster, work smarter, and maximize productivity across their existing AI stack.
We’re a fast-moving, early-stage team with a proven track record of building and scaling successful AI businesses. Freshly funded and growing, we’re looking for builders who want to help define the next frontier of human–AI collaboration.
Role Summary
You’ll own the 0→1 systems behind Base0’s Intelligence Layer—building the knowledge extraction, mapping, and retrieval systems that transform AI-native work.
This means experimenting with how to extract, represent, and retrieve knowledge from thousands of conversations—and turning those experiments into working systems that power real user experiences across our API and user-facing features/products.
If you’re excited by turning product ideas into working LLM systems and iterating through research, data, and prototypes to find what works, you’ll thrive here.
What You’ll Build
We’re solving one of the biggest challenges in modern AI workflows: fragmented context. Today, project knowledge is scattered across conversations, tools, and docs—forcing teams to spend more time steering AI than actually getting work done.
Base0 brings context, continuity, and observability to AI-powered work, enabling teams to ship faster, work smarter, and maximize productivity across their existing AI stack.
We’re a fast-moving, early-stage team with a proven track record of building and scaling successful AI businesses. Freshly funded and growing, we’re looking for builders who want to help define the next frontier of human–AI collaboration.
Role Summary
You’ll own the 0→1 systems behind Base0’s Intelligence Layer—building the knowledge extraction, mapping, and retrieval systems that transform AI-native work.
This means experimenting with how to extract, represent, and retrieve knowledge from thousands of conversations—and turning those experiments into working systems that power real user experiences across our API and user-facing features/products.
If you’re excited by turning product ideas into working LLM systems and iterating through research, data, and prototypes to find what works, you’ll thrive here.
What You’ll Build
- Core AI systems that extract and organize knowledge from AI conversations into structured, reusable context.
- Retrieval and memory infrastructure (GraphRAG vector search) that delivers precise, low-latency context to user workflows.
- Agentic systems that reason across stored context—handling retrieval, synthesis, and evaluation tasks autonomously.
- Prompt and orchestration frameworks that connect multiple models, tools, and data sources into end-to-end reasoning pipelines.
- Evaluation and telemetry systems to benchmark retrieval quality, latency, and overall intelligence performance.
- Fast prototypes to explore new product directions and validate user-facing capabilities.
- Strong Python engineering fundamentals—skilled in building performant, maintainable systems and services that connect data, models, and APIs.
- Deep understanding of retrieval architectures—embeddings, vector databases, hybrid or graph-based search, and caching strategies.
- Experience with LLM orchestration frameworks like LangChain, LlamaIndex, or custom-built agent systems.
- Proven ability to build and tune LLM-based agents for reasoning, synthesis, or evaluation tasks.
- Familiarity with prompt engineering and multi-step reasoning—designing structured flows that balance quality, latency, and cost.
- Exposure to fine-tuning or adapter training (LoRA, PEFT) and how to integrate tuned models into retrieval pipelines.
- Ability to work end-to-end—backend (FastAPI, Node) to quick front-end demos or dashboards for testing and iteration.
- Comfort operating in open-ended problem spaces, defining your own experiments, and driving them to working outcomes.
