GenAI Architect · 15+ years

Lenin S

AI Architect specializing in research and development — with an attitude: “There is no such thing as can’t.”

View experience Key achievements
Lenin S portrait
0 Years in AI & data

Profile & capabilities

End-to-end AI lifecycles — from strategy and architecture to production GenAI, governance, and cost-aware operations.

Agentic AI frameworks GenAI & RAG AI strategy & architecture MLOps / LLMOps AI governance Observability Token & cost optimisation Computer vision Machine & deep learning Data streaming Cloud & infrastructure NLP Delivery frameworks Project management
  • AI architecture — Scalable, maintainable AI systems aligned to business outcomes.
  • Machine learning & deep learning — Predictive models, neural networks, and advanced algorithms across data types.
  • Large language models — Fine-tuning for NLP and conversational AI.
  • NLP — Sentiment analysis, entity recognition, and text-centric products.
  • Computer vision — Object detection, image classification, and vision pipelines.
  • OCR — Document digitization and extraction workflows.
  • Distributed data — Hadoop, Spark, and large-scale processing patterns.
  • Visualization — Tableau, Power BI, and decision-ready reporting.
  • Decision science — Turning data into measurable business value.

Work experience

Progression from analytics and engineering to GenAI architecture and org-wide platforms.

GenAI Architect

Orion Innovation —Present
  • Define end-to-end AI lifecycles and deliver scalable GenAI/AI solutions — including RAG pipelines and conversational AI.
  • Lead research, prototyping, and productionisation paths with clear engineering standards.
  • Optimise infrastructure, model, and token economics without sacrificing quality.
  • Establish AI engineering best practices and mentor cross-functional teams.

Lead Data Scientist

Solverminds —
  • Develop and execute AI strategy aligned with organisational goals.
  • Partner with business stakeholders to frame problems and use data to propose actionable solutions.
  • Apply advanced algorithms and statistical methods across diverse data domains.
  • Run research programmes and build customer-facing and in-house prototypes and proofs of concept.
  • Identify high-value AI opportunities and integrate solutions across business units with measurable impact.
  • Lead design, development, and deployment of models; ensure robustness, scale, and maintainability.
  • Optimise cost across AI development and operations.
  • Communicate clearly — written and verbal — to align data needs and tell compelling outcomes stories.

Data Analyst

Mindtree —
  • Analytics, reporting, and stakeholder delivery in a global IT services context.

Test Engineer

Pyramid IT Consulting —
  • Quality engineering, test design, and release confidence for client systems.

Analyst

FBS Technologies —
  • Foundational analytics and business reporting roles.

Achievements & impact

Selected outcomes from enterprise AI, multi-agent systems, and production GenAI.

Credit approval AI

Intelligent multi-agent risk analysis for banking — external signals (news, competitors, financials) integrated into workflows. ~60% faster risk assessment and stronger decision quality. Architected cost-optimised LLM usage (~30% token savings).

Code modernisation agents

Multi-agent orchestration with LangGraph for large-scale modernisation — reverse and forward engineering at scale. Transformation timelines reduced from ~2 years to ~6 months (~75%).

Multi-agent SDLC platform

Centralised AI-powered SDLC platform for 200+ users — adopted across Orion Innovation for standardisation, automation, and efficiency.

Security & debugging agents

Enterprise-grade AI security and debugging agents — org-wide adoption, ~20% reduction in production defects, improved reliability.

Customer-service AI

Integrated AI into customer service — ~35% efficiency uplift. Low-latency RAG pipelines with sub–3s response targets.

Risk prediction engine

Advanced ML/DL for maritime risk — ~30% improvement in predictive accuracy.

Technical skills

Deep stack across GenAI, retrieval, vectors, MLOps, cloud, and secure delivery.

GenAI & agentic frameworks
LangChain, LangGraph, AutoGen, CrewAI, Semantic Kernel, LlamaIndex, MCP, A2A
LLM ecosystem
OpenAI / Azure OpenAI, Claude, Gemini, Hugging Face Transformers, vLLM, Ollama; fine-tuning with LoRA, QLoRA, PEFT
RAG & retrieval
RAG architecture, hybrid search, re-ranking, knowledge graph + RAG, chunking strategies, embeddings & evaluations
Vector & graph databases
Vector: Chroma, Pinecone, FAISS, Weaviate, Milvus, PGVector · Graph: Neo4j, Memgraph
MLOps / LLMOps
MLflow, Kubeflow, Weights & Biases, prompt versioning, Ragas, ARES, DeepEval, CI/CD for ML, guardrails
Observability & AI governance
Langfuse, Arize AI, responsible AI, LLM red/blue teaming
Security
AI security, prompt-injection handling, data privacy, RBAC
Cloud platforms
AWS Bedrock, SageMaker, Azure OpenAI, Azure AI Foundry, Vertex AI, BigQuery
DevOps & deployment
Docker, Kubernetes, Terraform, CI/CD pipelines
Programming & APIs
Python, R, SQL · FastAPI, Flask, Django, REST / GraphQL
Frontend
React, Streamlit

Certifications & education

Certificates

Education

B.E. — Electronics & Communication Engineering

Thangavelu Engineering College · Sep 2007 — May 2011

Let’s connect

Open to leadership architecture roles, advisory, and high-impact GenAI programmes.

Email Lenin LinkedIn