Hands-on AI Lead - Cyber Security Startup - CTC INR 75 L

CareerXperts Consulting; Pune District; On-site; INR 4,000,000 - 8,000,000; Full time

Job summary

CareerXperts Consulting is seeking an AI Team Lead to own the machine learning and agentic AI layer of our platform, and to lead a team of AI and ML engineers while staying hands-on in the code. You will set the technical direction, lead a team of AI and ML engineers, and stay hands-on in the code.

This is a player-coach role, not a management-only one. You will own local and self-hosted deployment for customers with private-cloud or sovereign constraints, and you will manage inference cost,

Qualifications

• 8+ years in ML, applied AI or data science with 4+ years deploying production systems.
• 2+ years leading or tech-leading an AI/ML team while remaining hands-on.
• Owned ML and agentic AI workflows in production with measurable impact.
• Strong Python, PyTorch/TensorFlow, SQL; familiarity with ML frameworks and APIs.

Skills

Machine learning

Production systems

Team leadership

LLM application engineering

Python

PyTorch / TensorFlow

SQL

MLOps

Tools

Docker

Kubernetes

OpenAI API

Hugging Face Transformers

LangChain

Job description

We are looking for an AI Team Lead to own the machine learning and agentic AI layer of our platform and lead the team building it. You will set the technical direction, lead a team of AI and ML engineers, and stay hands-on in the code. This is a player-coach role, not a management-only one. The role spans both halves of the problem. Classical machine learning does the ranking, classification and behavioural baselining that decides what deserves attention; LLM-driven agents do the reasoning and investigation on top. You need genuine depth in both, and judgement about which one a given problem actually calls for. Ownership here is real. You decide which models run where and why, you run the R&D that answers that question, you own local and self-hosted deployment for customers with private-cloud or sovereign constraints, and you own how inference cost, latency and quality hold up as the product scales.

Required Skills

• 8+ years in machine learning, applied AI or data science, with at least 4 years deploying and owning production systems.
• 2+ years leading or tech-leading an AI/ML team while remaining hands-on. You have set direction, reviewed others' work, and shipped your own.
• You have owned ML and agentic AI workflows in production — chose the approach, shipped it, and stayed accountable for how it behaved. Research and prototypes are not the same thing.
• Strong knowledge of machine learning algorithms for classification, ranking, clustering and anomaly detection on structured and semi-structured data such aslogs, alerts, identity and cloud events.
• Strong LLM application engineering: retrieval-augmented generation and grounding, agent orchestration and tool use, structured output, prompt engineering, and fine-tuning or domain adaptation.
• Hands-on with local and self-hosted models (vLLM, TGI, Ollama or equivalent):quantisation, GPU and hardware sizing, and the throughput, latency and cost trade-offs that come with them.
• Experience building evaluation harnesses — you have measured whether a model, prompt or agent change actually made things better, and can explain how you knew.
• Strong Python with PyTorch, TensorFlow or scikit-learn; proficiency in SQL; and familiarity with LLM frameworks and APIs (OpenAI, Hugging Face Transformers, LangChain, LlamaIndex).
• MLOps in production — CI/CD for ML, versioning, monitoring and observability — on AWS, GCP or Azure with Docker and Kubernetes.
• Strong written and verbal communication skills.
• Comfortable working in a startup environment with high ownership.

Preferred Skills

• Nice to have
• Security-specific ML: UEBA, entity behaviour baselining, graph analytics.
• Working knowledge of security operations and MITRE ATT&CK or ATLAS.
• Experience with multi-tenant, private-cloud or air-gapped deployment and the isolation constraints that come with it.
• Feedback-driven systems where analyst input improves future predictions.
• Publications, patents or open-source contributions in ML/AI.