Senior Software Engineer — AI Agentic & RAG (Hybrid)

Socket.dev; Boston (MA); Hybrid; USD 110,000 - 200,000; Full time

Job summary

Accordion Intelligence Lab is hiring engineers to build real agentic systems, including multi-agent components and RAG pipelines for PE-backed engagements. You’ll work under Staff-level engineers on live client engagements writing agent logic, RAG pipelines, and eval harnesses.

This highly technical, hands-on role demands fast delivery, learning on the job, and clear communication with non-technical stakeholders. Remote-ready in a hybrid, client-facing environment with rapid pacing.

Qualifications

• Solid Python and/or TypeScript/React fundamentals — you write clean code and can reason about what a framework is doing underneath
• Some hands-on exposure to LLM application development: RAG, tool use, prompt engineering, or agent orchestration
• Genuine use of AI coding tools (Claude Code, Cursor, or similar) in your own workflow
• Comfort with ambiguity and a fast pace — you won’t always get a clean spec, and that’s the job
• Willingness and ability to learn quickly in a client-facing, consulting-style environment

Responsibilities

• Build and ship components of multi-agent systems and RAG pipelines for PE-backed portfolio company engagements, under the guidance of more senior engineers
• Contribute to prompt engineering, tool design, and deployment for production AI systems
• Write and run evals as part of eval-driven development — define success criteria before you build, not after
• Help translate client requirements into technical tasks, with support from your pod on the ambiguous parts
• Build observability into the systems you ship so quality is measurable, not assumed
• At the Senior level: independently own a workstream end-to-end and present your architecture decisions to internal and client stakeholders

Skills

Python

TypeScript/React

LLM application development

AI coding tools

Tools

LangChain/LangGraph

AutoGen

DSPy

Job description

Accordion Intelligence Lab is hiring engineers to build real agentic systems, including multi-agent components and RAG pipelines for PE-backed engagements. You’ll work under Staff-level engineers on live client engagements writing agent logic, RAG pipelines, and eval harnesses.

This highly technical, hands-on role demands fast delivery, learning on the job, and clear communication with non-technical stakeholders. Remote-ready in a hybrid, client-facing environment with rapid pacing.