Asks for

The ad says“Deep understanding of retrieval architectures—embeddings, vector databases, hybrid or graph-based search, and caching strategies.”
The ad says“Deep understanding of retrieval architectures—embeddings, vector databases, hybrid or graph-based search, and caching strategies.”
The ad says“Ability to work end-to-end—backend (FastAPI, Node) to quick front-end demos or dashboards for testing and iteration.”
The ad says“Exposure to fine-tuning or adapter training (LoRA, PEFT) and how to integrate tuned models into retrieval pipelines.”
The ad says“Deep understanding of retrieval architectures—embeddings, vector databases, hybrid or graph-based search, and caching strategies.”
The ad says“Experience with LLM orchestration frameworks like LangChain, LlamaIndex, or custom-built agent systems.”
The ad says“Experience with LLM orchestration frameworks like LangChain, LlamaIndex, or custom-built agent systems.”
The ad says“Exposure to fine-tuning or adapter training (LoRA, PEFT) and how to integrate tuned models into retrieval pipelines.”
The ad says“Familiarity with prompt engineering and multi-step reasoning—designing structured flows that balance quality, latency, and cost.”
The ad says“Ability to work end-to-end—backend (FastAPI, Node) to quick front-end demos or dashboards for testing and iteration.”
The ad says“Exposure to fine-tuning or adapter training (LoRA, PEFT) and how to integrate tuned models into retrieval pipelines.”
The ad says“Familiarity with prompt engineering and multi-step reasoning—designing structured flows that balance quality, latency, and cost.”
The ad says“Strong Python engineering fundamentals—skilled in building performant, maintainable systems and services that connect data, models, and APIs.”
The ad says“Deep understanding of retrieval architectures—embeddings, vector databases, hybrid or graph-based search, and caching strategies.”

About Symmetry

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.

Symmetry 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 Symmetry’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.

Skills We’re Looking For

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.

If you thrive on autonomy, clarity, and collaboration and want to build the connective tissue between humans and AI systems, Symmetry is where you’ll do your best work.