AI​/ML Engineer

From bid to build to paid. From the office to the field to the entire company. Sledge is Construction AI.Sledge is the agentic, AI-native operating system for construction, handling the busywork of running a construction business so builders can spend more time building and less time on paperwork. Built for subcontractors and small to mid-size general contractors, Sledge is already live with a growing group of beta users.

The AI isn't a feature bolted onto Sledge, it's the core of the product.

Sledge is hiring an ML/AI Engineer to build and run the AI systems that make Sledge Construction AI, not just construction software with a chatbot bolted on. Sledge already runs a multi-agent AI system in Python and FastAPI, with Pinecone for vector storage, and an internal agent fleet that includes agents like Forge, Mimir, Anvil, Mason, Beacon, Scout, Warden, Crane, and Vault.

This role reports to Sledge's engineering team and works hands-on inside that system, not around it.

This is a build role, not a research role. Sledge needs agents that work reliably in production, on a weekly deployment cadence, not a paper or a demo.

Build, extend, and maintain Sledge's multi-agent AI system, built in Python and FastAPIDesign and improve retrieval and vector search using Pinecone Build and refine the agents in Sledge's internal fleet, and design new ones as Sledge needs them Write and tune prompts, and build the evaluation loops that catch when an agent gets worse, not just better

Fine-tune or train models when prompting and RAG aren't enough to get the job done

Work with the rest of engineering to ship AI features on Sledge's weekly deployment cadence

Work with QA to get AI features properly tested, including through Sledge's Playwright-based testing tool Debug and improve agents in production, not just in a notebook

Must-haves

Required:

hands-on production experience building with LLMs, not just personal projects or coursework

Required:

strong Python skills, comfortable in a FastAPI-style backend

Required:

hands-on experience building RAG pipelines, from chunking and embeddings through retrieval and reranking

Required:

experience training or fine-tuning models, not just prompting a hosted APIReal experience with vector databases beyond knowing the term: indexing strategies, embedding models, and retrieval quality, ideally including Pinecone Experience building or working with multi-agent systems, including agent-to-agent orchestration and handoffs, not just a single agent calling tools

Comfortable owning an AI feature from prototype to production, not just the model itself

Comfortable moving at startup speed, shipping weekly

Nice-to-haves

Experience with agent orchestration approaches or frameworks, and a clear opinion on when to use one versus build custom

Familiarity with LLM evaluation techniques: evals, hallucination detection, and prompt regression testingA habit of following LLM and agent research and turning it into something that ships, not just reading it Experience with Type Script, Next.js, or Express Experience with Supabase, PostgreSQL, or Drizzle ORMExperience with AWS Interest or experience in construction, trades, or B2B software

Fully remote, work from anywhere in the world

Flexible hours, work when you're most productive

Direct access to the founding team and engineering, no layers of management

You're building the actual AI core of the product, not a wrapper around someone else's model

Ship weekly, see your work in production fast

Sledge is an equal opportunity employer. Sledge welcomes applicants of every background and does not discriminate on the basis of race, color, religion, sex, national origin, age, disability, or any other protected status.