Agentic AI Engineer

FlytBase; Pune District; On-site; INR 3,000,000 - 6,000,000; Full time

Benefits offered by this job

Competitive pay

Flexible work hours

Unlimited PTO

Pune HQ presence

Job summary

FlytBase is seeking an Agentic AI Engineer to design multi-agent frameworks where AI not only suggests but acts. You’ll build autonomous decision-making systems that control drone fleets in real-world missions and push AI to operate at scale.

This role blends advanced ML engineering with practical deployment, emphasizing safety, observability, and low-latency performance on AWS/GCP. Pune-based, you’ll contribute to a fast-moving, high-velocity team shipping production code.

Qualifications

• Engineers with proven AI project experience, especially multi-agent systems, RAG, or decision automation.
• Proficiency in Python, PyTorch/TensorFlow, Hugging Face, and cloud environments (AWS/GCP).
• Familiarity with vector databases (Pinecone, Milvus) and agentic frameworks (LangChain, CrewAI, LangGraph, AutoGen).
• Thrives in ambiguity, learns fast, and values speed without compromising safety.

Responsibilities

• Architect multi-agent frameworks for autonomous decision-making and agent communication.
• Implement RAG pipelines and persistent memory to ground agents in domain knowledge.
• Develop Vision-Language Model (VLM) solutions for aerial and sensor imagery analysis.
• Enable agents to autonomously execute Python functions, API calls, and data analysis.
• Deploy AI workflows on AWS/GCP with GPU efficiency and latency optimizations.
• Implement observability and debugging tools (e.g., LangSmith) to monitor agent behavior.
• Design human-in-the-loop safety mechanisms, including fallback protocols and bias checks.

Skills

Python

PyTorch/TensorFlow

Hugging Face

AWS/GCP

LangChain

LangGraph

AutoGen

RAG pipelines

Multi-agent systems

Decision automation

Tools

Pinecone

Milvus

CrewAI

LangChain

AutoGen

Job description

Architect and deploy Agentic AI systems that enable autonomous drones to perceive, reason, and act independently—bridging cutting-edge AI with real-world robotics at global scale.

Software Development

Pune, India

Full-time

0+ Years

What We Do

FlytBase is an AI-native company building the physical AI layer for the sky. Its platform powers fully autonomous drone-in-a-box operations across energy, utilities, security, and infrastructure sites worldwide—integrating drones, docks, edge AI, and enterprise workflows into one intelligent system. Headquartered in California with a fast-moving engineering base in Pune, FlytBase is redefining how the physical world is monitored and managed. The company operates at AI speed, with autonomy and intelligence at its core. At FlytBase, teams work in a high-agency, high-velocity environment where AI isn’t a side project—it’s the operating system for how they think, build, and grow. With 300+ global deployments and 50+ partners in our ecosystem, FlytBase is powering the shift from human-operated to self-driven skies. Our platform transforms off-the-shelf drones into intelligent systems that sync effortlessly with enterprise infrastructure.

The Role

As an Agentic AI Engineer, you’ll design multi-agent AI frameworks where AI doesn’t just make suggestions—it takes action. You’ll integrate reasoning, perception, and execution into autonomous workflows that directly control drone fleets in the real world. This role blends deep AI/ML engineering with practical deployment skills to build scalable, safe, and high-performance systems.

What You'll Own

• Architecting multi-agent frameworks for autonomous decision-making and agent-to-agent communication.
• Implementing Retrieval-Augmented Generation (RAG) pipelines and persistent memory to ground agents in domain-specific knowledge.
• Developing Vision-Language Model (VLM) solutions for aerial and sensor imagery analysis.
• Enabling agents to autonomously execute Python functions, API calls, and data analysis.
• Deploying AI workflows on AWS/GCP with optimizations for GPU efficiency and latency.
• Implementing observability and debugging tools (e.g., LangSmith) to monitor and refine agent behavior.
• Designing human-in-the-loop safety mechanisms, including fallback protocols and bias checks.

Best Suited For

• Engineers with a proven track record of AI projects—especially multi-agent systems, RAG implementations, or decision automation.
• Proficiency in Python, PyTorch/TensorFlow, Hugging Face, and cloud environments (AWS/GCP).
• Familiarity with vector databases (Pinecone, Milvus) and agentic frameworks (LangChain, CrewAI, LangGraph, AutoGen).
• Builders who thrive in ambiguity, learn fast, and value speed without compromising safety.

What Makes You a Flyter

• You use AI tools as your co-engineer
• You move fast without breaking reliability
• You jump stacks instead of staying in your lane
• You thrive in messy, ambiguous, high-impact problem spaces
• You want your code running on fleets of autonomous robots
• You bias toward action and clarity, not waiting for perfection

What We Don't Want

• Engineers who only build demos but can’t ship production code.
• Specialists who refuse to work outside their comfort zone.
• People who follow AI hype without grounding solutions in practical use cases.
• Anyone who needs constant direction instead of driving their own roadmap.

Why FlytBase?

Here, your AI won’t just run in benchmarks—it will control fleets of autonomous drones in real missions worldwide. You’ll join a high-velocity, high-trust team building a first-of-its-kind technology category with real impact.

The Perks

• Competitive pay and performance bonuses.
• AI-first development environment with cutting-edge tools.
• Flexible work hours with core collaboration at Pune HQ.
• Unlimited paid time offs.
• Direct influence on a product deployed in high-stakes, global environments.