Location: Noida (In office)
Company: Zynence Labs Pvt. Ltd.
Experience: 0–2 years
Employment Type: Full-time
About Zynence
Zynence is building an AI-native, full-lifecycle hardware engineering platform, a connected workspace that takes a hardware product from requirements and system architecture through BOM/cost, firmware, schematic/PCB, simulation, validation, and manufacturing. The platform runs on eleven purpose-built AI agents and a RAG-grounded knowledge layer, with a core architectural bet that generative AI handles drafting and structuring while deterministic engines compute anything an engineer would sign off on as a number, cost rollups, component derating, trace current capacity, battery runtime. We're targeting the sectors where design traceability is non-negotiable: medtech, aerospace & defense, robotics, EV hardware, and industrial IoT.
Role Overview
We're looking for an AI/ML Engineer who is equally comfortable building and shipping full-stack web features and building the AI/ML systems that power them. You'll work across our agent architecture and knowledge layer,
from retrieval pipelines and LLM-powered agents to the web application and cloud infrastructure that put them in front of users. This is a hands-on, build-things role at an early-stage product company: you'll ship real features that go into a live platform, not a research sandbox.
Key Responsibilities
AI/ML & Agent Development
Design, build, and iterate on LLM-powered agents within our multi-agent architecture (AGT-001–AGT-011).
Build and improve retrieval-augmented generation (RAG) pipelines for our knowledge layer (AskZynence), including embedding generation, vector search, and retrieval quality tuning.
Work within our generative/deterministic split — know when a task belongs to a language model and when it belongs to a deterministic engine, and build accordingly.
Evaluate and integrate LLM APIs (OpenAI, Anthropic, or similar), including prompt design, structured output handling, and cost
📌 Full Stack AI Engineer (Noida)
🏢 Zynence
📍 Noida