Founding Engineer

About Manufex Manufex is a managed B2B marketplace for custom manufacturing procurement. We connect US hardware teams in aerospace, medtech, robotics, and defense with a vetted supplier network, and we run the entire procurement loop (sourcing, DFM, quoting, compliance, quality, and fulfillment) through a coordinated stack of AI agents. Our wedge is radical pricing transparency: itemized, defensible quotes instead of the opaque markups the incumbents are built on. We're pre-launch and building fast.

About the Role We're looking for a hands-on core engineer who can own the entire platform (from backend services and the agent stack to the customer-facing frontend) and help shape product and technical direction from the ground up. This is not a role for someone who wants to manage; it's for someone who wants to build. You'll be one of the first engineers, with outsized ownership over architecture, product decisions, and engineering design, from HLD to LLD to implementation. You'll define the patterns the rest of the team will follow.

What You'll Do

• Build the core marketplace: buyer, supplier, and admin surfaces, matching, and the transaction and fulfillment backbone.
• Build and ship the agent stack that runs sourcing, DFM, quoting, negotiation, compliance, and order orchestration between buyers and suppliers: production agents, not demos.
• Own the full stack: FastAPI services, an event-driven backend, and Next.js frontends.
• Extend the quoting engine: CAD/geometry ingestion (STEP), parametric cost models, and DFM reasoning that turn a part file into a transparent, itemized price.
• Design the context and prompt strategies, multi-model routing with provider fallback, and MCP tool integrations that make agents reliable and cost-aware in production.
• Stand up evaluation, observability, and guardrails for agentic workflows: calibrationgated confidence, LLM-as-judge evals, and audit trails.
• Build and defend the privacy architecture that keeps buyers and suppliers anonymized by default, with controlled de-anonymization on admin-only paths.
• Make foundational architecture calls and work directly with the founders on direction, priorities, and tradeoffs.

What We're Looking For Core Engineering (required)

• 0-to-1 experience is mandatory. You've taken a product from nothing to shipped in production, ideally at an early-stage startup, not just maintained or scaled an existing one.
• 6+ years building and shipping production software, with real end-to-end ownership of systems.
• Deep proficiency in Python (we build services and agents on FastAPI) and TypeScript / React (we build the buyer, supplier, and admin surfaces on Next.js).
• Strong data-modeling instincts and production experience with a relational database (PostgreSQL): schema design, normalization tradeoffs, indexing, and query performance.
• Comfort with event-driven and async architectures: event streaming (Kafka), message queues, and saga/outbox patterns.
• Strong systems fundamentals: API design, distributed systems, performance and scale tradeoffs, and a security-first default.
• Hands-on experience building and owning CI/CD pipelines, infrastructure-as-code, and cloud infrastructure (AWS or GCP), ideally as an early or sole engineer.

AI Engineering (required)

• 2+ years hands-on AI engineering: building and shipping LLM-powered features in production, not just experimentation.
• 1+ years specifically with AI agent workflows and orchestration: agents that plan, call tools, use memory, and operate semi-autonomously within guardrails.
• Context engineering and prompt engineering as production disciplines: structuring, compressing, and budgeting context windows for reliability, cost, and accuracy. Familiarity with several of the following (or equivalent):
• Agent orchestration and state management (LangGraph or similar)
• MCP (Model Context Protocol) for tool and data integration with agents
• Multi-model / multi-provider setups with fallback (we run Gemini 2.5 Pro/Flash and Claude via OpenRouter, with an OpenAI fallback)
• Vector databases and retrieval pipelines (RAG): chunking, embeddings, hybrid search, reranking
• Eval frameworks for LLM output quality: LLM-as-judge, benchmark-style testing, calibration

Mindset

• Comfortable with ambiguity: this is a 0-to-1 environment with evolving requirements.
• Bias toward shipping and iterating over long design cycles: fast, but not at the expense of quality and reliability.
• Genuine interest in agentic AI as a product surface, not just a backend tool.
• Curiosity about the physical world: parts, processes, and how things actually get made.

Nice to Have

• Experience in marketplace, fintech, or B2B transactional platforms.
• Exposure to manufacturing, supply chain, hardware, or procurement. Bonus for anyone who's touched CAD, CNC / sheet-metal / additive processes, DFM, or quoting.
• Experience with e-signature, logistics/shipping APIs, or compliance workflows (export-control awareness is a plus given our verticals).
• Experience deploying and operating Next.js apps on Vercel.

What You Get

• Founding-level equity, competitive compensation, and genuine ownership over a category-defining product.
• Direct influence on architecture, tech stack, and hiring as the team grows.
• A short line from part file to shipped feature, working directly with a founding team that has deep manufacturing-marketplace operating experience (including ex-Xometry).
• High-autonomy environment, working out of our Bengaluru office.

Compensation Founding-level equity, a competitive salary, and full transparency on the cap table before you sign. We are pre-launch and early, with no revenue yet. Transparency is our product, so we will be straight with you about the stage and the risk.

How to apply Send a short note about yourself and one system you took from zero to production: what you owned, the hardest technical tradeoff you made, and what you would do differently now, in our DM . You can also reach out on

careers@manufex.io