Experience
- At least 4 years
- The applicants should have experience in the following business area(s): Artificial Intelligence (AI) Startup, IT Enabled Service
Education
- Bachelor of Science (BSc) in Computer Science & Engineering
Requirements
- At least 4 years
- The applicants should have experience in the following business area(s): Artificial Intelligence (AI) Startup, IT Enabled Service
Skills Required
Additional Requirements
4+ years in software engineering, architecture, or applied AI OR equivalent experience (including a non-traditional path into the field).
Substantial hands-on experience with LLM-based, RAG and agentic systems in production, not only in prototypes, OR equivalent experience.
Deep knowledge of architecture across distributed systems and applied AI, with hands-on build ability.
Judgment about risk and autonomy boundaries.
Explaining design decisions to engineers, executives and auditors.
Reliability and containment design for autonomous systems.
Agent frameworks, tool-calling protocols and multi-agent coordination.
Evaluation methodology for agentic systems.
Security for autonomous systems: prompt injection, guardrails and sandboxing.
Deployment in regulated industries.
Responsibilities & Context
Define the reference architecture for agentic systems across products: orchestration, tool interfaces, delegation and human oversight.
Design and architect the agent's memory and knowledge systems, including short-term/context memory, long-term persistent memory, RAG pipelines, retrieval strategies, memory summarization, and context management to ensure agents retain and retrieve relevant information accurately and efficiently.
Set standards for autonomy boundaries — what an agent may do unsupervised and what requires approval — and ensure they are enforced technically.
Design the end-to-end evaluation and assurance framework for agentic AI systems, including pre-deployment qualification and continuous post-deployment monitoring to detect performance drift, regressions, safety issues, and emerging failure modes in regulated or high-consequence environments.
Architect client solutions in the most demanding engagements, and prototype the highest-risk components directly.
Work with the research team to bring new agentic techniques into the platform, and identify which are not yet ready.
Establish reliability and containment patterns: failure handling, rollback, idempotency and prevention of compounding agent errors.
Mentor engineers and raise collective judgment in designing agentic systems.
Job Other Benifits
- T/A,Provident fund,Mobile bill,Weekly 2 holidays,Insurance,Gratuity
- Festival Bonus: 2
- Salary Review: Yearly
- Lunch Facilities: Partially Subsidize
As per company policy