We are expanding our AI products portfolio and looking for a Lead Software Engineer—AI Products to take ownership of core engineering work across our cross‑product AI cluster. This role is ideal for someone who enjoys designing and shipping production‑grade LLM and agentic systems, working across multiple SaaS products, and setting strong technical standards while staying hands‑on.
What We’re Looking For:
Engineering Experience:
7+ years of commercial software engineering experience.
Strong backend engineering background with experience in multi‑tenant SaaS systems (tenant isolation, scalability, reliability, security basics).
Hands‑on engineer who codes daily in TypeScript/Node.js.
Comfortable working as a full‑stack engineer — primarily backend‑focused, but able to implement and review frontend features using modern JavaScript frameworks (e.g., React.js, Vue.js, or similar).
Proven track record of delivering production systems end‑to‑end, not just prototypes or PoCs.
AI, LLMs & Agentic Systems. Recent (last 12 months) hands‑on experience shipping LLM‑based or agentic features into production, for example:
Retrieval‑augmented generation (RAG) over a vector store,
Tool‑using agents or copilots integrated into an application,
MCP (Model Context Protocol) server development,
Multi‑step agent workflows that call internal and external services.
Familiarity with agent and orchestration frameworks, such as Claude Agent SDK, OpenAI Agents, or Langgraph—and ability to discuss trade‑offs between approaches.
Agentic coding mindset. Experience working with Claude Code or OpenAI Codex or Google Antigravity or Cursor, using specs (e.g., Spec‑Driven Development / SDD or similar approaches) to design, implement, and evolve agentic systems, rather than just calling a single LLM endpoint.
Prompt management and observability (e.g., Langfuse or equivalent tools for tracing prompts, tool calls, and failures).
Model gateways or routing layers (for example, LiteLLM or LangChain or Vercel’s AI SDK, or a clearly comparable framework) to work with multiple models/providers behind a common interface—understanding the pattern matters more than a specific library.
PostgreSQL plus vector extensions such as pgvector (or equivalent) for storing and querying embeddings.
Exposure to agentic workflow setup and runtime governance in production: defining how agents are scheduled, executed, and monitored (e.g., background job runners such as Temporal, Trigger.dev, Hatchet, or similar) and applying quality gates (evaluations, safety checks, and review processes) around agent behavior.
Collaboration & Communication
Comfortable taking technical lead responsibility on complex features: making design decisions, defining quality gates (reviews, tests, evaluations), documenting them, and driving them to completion with the team.
Experience writing clear technical decision records (ADR‑style documents or similar) and discussing options and trade‑offs with product stakeholders.
Ability to work closely with product owners and senior leadership, translating product ideas into realistic, implementable plans.
Nice to Have
Awareness of the Salesforce ecosystem
Python experience is a strong plus given the AI ecosystem
Experience building or integrating MCP servers.
Domain familiarity with commercial real estate (CRE) or broader real estate / proptech domains.
Experience designing trust/provenance/evidence layers for AI outputs (citations, evidence tracking, verification workflows).
Hands‑on experience with headless browser / rendering stacks (e.g., Headless Chrome, HTML5 canvas, or similar approaches for generating visual artifacts and documents).
Background working in product‑led organizations, collaborating closely with product and design.
What You’ll Do:
Build and Evolve AI Products. Design and implement LLM‑ and agent‑powered features across the ACX AI Products portfolio, including the following:
A document‑intelligence
