Threat Research Developer (AI/ML)

Job Title - Threat Research Engineer / Developer

Job Location - Pune, Maharashtra

MUST Have Skills - Python coding, Technical research, Analysing AI Agents, Guardrails for AI Bots

Our client is looking for a highly analytical and hands-on Threat Research Engineer to help build their next-generation AI-driven Agent Observability + Trust Platform.

In this role, you will work on identifying and analyzing behavioral patterns across various types of AI agents using machine learning, telemetry analysis, and experimental pipelines. You will help design and operate research environments that ingest agent logs, execution traces, prompts, metadata, and interaction patterns to detect anomalies, risks, and suspicious behaviors.

Key Responsibilities

• Analyze behavioral patterns across AI agents and autonomous workflows
• Identify anomalies, suspicious execution flows, and emerging threat indicators
• Study telemetry, prompts, traces, logs, and metadata generated by AI systems
• Research patterns related to:

Agent misuse

Prompt manipulation

Unsafe execution paths

Hallucination and drift behavior

Multi-agent interaction risks

• Design and maintain isolated research/lab environments
• Build pipelines to process:

Agent logs, Conversation data

Execution traces

Metadata streams

• Simulate agent behaviors and attack scenarios for analysis

Required Skills

• 3–5 years of experience in software engineering, product analysis, or technical research roles
• Strong hands-on experience with Python
• Solid understanding of:

AI/ML engineering

Security analytics

Data engineering

Threat research

• Strong analytical and problem-solving skills
• Ability to break down complex systems into understandable components

Preferred Skills (Nice to Have)

• Experience with:

OpenAI / Anthropic / open-source LLMs

Vector databases & embeddings (Pinecone, Weaviate, FAISS, etc.)

Prompt engineering and evaluation frameworks

• Exposure to:

AI observability / governance platforms

SaaS product ecosystems

Who Should Apply

• Engineers curious about how AI products are built internally
• Individuals who enjoy breaking down systems and understanding how things work
• Enjoys analyzing patterns and systems
• Likes experimentation and research-driven engineering