InLustro
Website:
inlustro.co
Job details:
Company Description
InLustro is revolutionizing workforce readiness through AI-driven Job Simulations that replicate real work scenarios, enabling organizations to hire, train, and scale talent with accuracy and efficiency. Since 2020, we have partnered with over 60 institutions and 50 enterprises, impacting more than 20,000 learners by providing tailored solutions for IT, BFSI, engineering, and business roles. Our platform emphasizes evidence-based assessments, skill verification, and experiential learning to support hiring, onboarding, and training processes. Headquartered in Chennai and guided by leading global investors, including Accelerating Asia Ventures (Singapore), InLustro aims to be the global standard for employability with national accolades such as the Atma Nirbhar Bharat Award and the Young Achiever Award.
Role Overview:
You will design and build production-grade AI systems that operate inside structured, stateful environments, powering simulations where users (and AI agents) must execute real workflows under constraints.
Your work will sit at the intersection of:
- LLM systems
- decision engines
- workflow orchestration
- tool-integrated environments
You will be responsible for building systems where AI acts, decides, and is evaluated under real-world conditions.
What You Will Build:
- Agentic systems that operate within constrained simulation environments (not open-ended prompts)
- LLM pipelines with grounding, verification, and policy enforcement layers
- Decision engines combining rules, probabilistic reasoning, and model outputs
- Simulation backends (state machines, event logs, replay systems)
- Evaluation systems that score execution based on telemetry, not just outputs
- Tool-integrated workflows (APIs, enterprise systems, sandboxed actions)
Core Responsibilities:
- Design and implement LLM-driven systems with deterministic constraints
- Build multi-step reasoning pipelines (planning → grounding → execution → verification)
- Develop stateful simulation architectures (not stateless inference)
- Integrate AI models with real or emulated tools/APIs
- Create evaluation frameworks to measure decision quality and execution correctness
- Optimize systems for latency, reliability, and reproducibility
- Work closely with product and domain teams to convert workflows into executable systems
Required Qualifications:
Strong foundation in computer science, data structures, and systems design
Experience building end-to-end AI systems in production (not just notebooks/models)
Hands-on expertise with python (must)
At least one ML/LLM framework (e.g., PyTorch, TensorFlow)
Experience with LLM application patterns:
- RAG systems
- Tool use / function calling
Prompt + system design
Understanding of how to control model behavior (not just train models)
Ability to design systems that are robust to edge cases and failure modes
Strong Signals (Highly Preferred):
- Experience with agent frameworks / multi-agent systems
- Familiarity with orchestration tools (e.g., Temporal, Airflow, custom pipelines)
- Experience building evaluation or benchmarking systems for AI
- Exposure to production infra (APIs, databases, distributed systems)
- Understanding of state machines / workflow engines
- Prior work on AI in high-stakes environments (finance, ops, healthcare, etc.)
What This Role Is NOT:
- Not a pure model training / research role
- Not limited to data analysis or dashboarding
- Not prompt tinkering without system-level ownership
Click on Apply to know more.