We are seeking an accomplished AI Architect with 15+ years of technology experience to define and deliver enterprise-scale AI solutions spanning machine learning, Generative AI, Agentic AI, engineering intelligence, and CAD drawing understanding. The role combines architecture leadership with hands-on technical depth to convert complex business and engineering problems into secure, scalable, production-ready solutions across cloud, on-premises, hybrid, and edge environments.

Role mandate: Own the journey from business problem and engineering data to AI architecture, PoC, industrialization, governance, and measurable value.

Key Responsibilities

- Define end-to-end enterprise architectures for AI/ML, Generative AI, Agentic AI, computer vision, multimodal AI, NLP, forecasting, and predictive analytics.

- Architect agentic AI solutions with planners, specialized agents, tool use, memory, orchestration, human-in-the-loop controls, guardrails, and observability.

- Design AI pipelines for CAD and engineering drawings, including 2D/3D CAD, P&IDs;, schematics, layouts, technical diagrams, and associated metadata.

- Lead approaches for drawing ingestion, raster/vector processing, OCR, symbol detection, object and relationship extraction, drawing classification, redline/change detection, and engineering document retrieval.

- Integrate AI with CAD/PLM/BIM/GIS/document-management ecosystems and engineering formats such as DWG, DXF, DGN, IFC, STEP, IGES, PDF, and image-based drawings, where applicable.

- Design RAG and knowledge-driven solutions using embeddings, vector databases, graph databases, knowledge graphs, semantic search, and enterprise content repositories.

- Evaluate and select models, vision-language models, frameworks, model-serving platforms, vector stores, graph technologies, and cloud or on-premises AI services.

- Define scalable AI platforms covering ingestion, data preparation, feature/model development, APIs, inference, workflow orchestration, monitoring, and enterprise integration.

- Establish MLOps and LLMOps practices covering experiment tracking, evaluation, model and prompt versioning, deployment, monitoring, drift, feedback, and governance.

- Provide guidance on GPU infrastructure, edge AI, model compression, quantization, pruning, inference optimization, latency, throughput, and cost.

- Ensure security, privacy, responsible AI, intellectual-property protection, explainability, reliability, and regulatory compliance are embedded by design.

- Lead architecture reviews, customer workshops, technology evaluations, PoCs, proposal solutioning, estimation, and technical defense discussions.

- Mentor AI engineers, data scientists, ML engineers, software engineers, and solution architects across multiple delivery teams.

Core Architecture Domains

Domain

Expected Capability

Generative & Agentic AI

LLMs, RAG, multimodal AI, autonomous and semi-autonomous agents, multi-agent orchestration, tool integration, memory, evaluation, guardrails, and human oversight.

CAD & Engineering AI

CAD and drawing understanding, OCR,



vision-language models, geometry and symbol extraction, BOM/tag linkage, P&ID; intelligence, change detection, drawing search, and engineering knowledge graphs.

AI/ML & Analytics

Machine learning, deep learning, computer vision, NLP, anomaly detection, time-series forecasting, predictive maintenance, and optimization.

AI Platforms

Data ingestion, model development, feature pipelines, API/microservice architecture, model serving, GPU platforms, observability, and enterprise integration.

MLOps / LLMOps

CI/CD, experiment tracking, registries, evaluation, prompt/model lifecycle, monitoring, drift detection, feedback loops, and governance.

Enterprise Architecture

Cloud, on-premises, hybrid and edge patterns; security; integration; scalability; resilience; cost optimization; and reusable reference architectures.

Required Technical Skills

AI, GenAI & Agentic AI

- Python; PyTorch and/or TensorFlow; Hugging Face ecosystem

- LLMs, embeddings, RAG, multimodal and vision-language models

- Agent frameworks, workflow orchestration, tool/function calling, memory and evaluation

- Fine-tuning, prompting, model evaluation, safety, guardrails and optimized inference

CAD, Drawings & Engineering Data

- Hands-on understanding of CAD/engineering drawing structures, layers, blocks, symbols, geometry and metadata

- Experience with OCR, document AI, computer vision and VLM-based extraction from complex technical drawings

- Integration patterns for CAD/PLM/BIM/GIS and engineering document repositories

- Familiarity with common formats such as DWG, DXF, DGN, IFC, STEP, IGES, PDF and raster images

Data, Cloud & Platforms

- One or more of Azure, AWS or Google Cloud; Databricks or Microsoft Fabric is advantageous

- Vector databases, graph databases/knowledge graphs, relational/NoSQL and time-series platforms

- Docker, Kubernetes, APIs, microservices, event-driven integration and GPU infrastructure

- Hybrid, on-premises and edge deployment architectures

MLOps, LLMOps & Governance

- CI/CD for AI, experiment tracking, model registry, prompt/model versioning and automated deployment

- Quality evaluation, telemetry, model monitoring, hallucination and drift detection, and feedback management

- Responsible AI, privacy, cybersecurity, access control, lineage, auditability and model risk governance

Experience & Qualifications

- 15+ years of overall experience in software, data, cloud, engineering systems, analytics, or AI-enabled transformation.

- At least 5 years of significant AI/ML, advanced analytics, or intelligent automation experience, including architecture ownership.





- Proven delivery of enterprise-scale AI/ML or Generative AI solutions in production.

- Demonstrated experience designing Agentic AI, RAG, multimodal, computer vision, or knowledge-graph solutions.

- Experience solving use cases involving CAD drawings, P&IDs;, schematics, engineering documents, asset information, or industrial data is strongly preferred.

- Solid understanding of enterprise integration, APIs, data architecture, security, cloud and distributed systems.

- Bachelor’s or Master’s degree in Computer Science, Engineering, AI/ML, Data Science, or a related discipline.

Leadership & Consulting Expectations

- Translate ambiguous business and engineering challenges into a prioritized AI roadmap and defensible target architecture.

- Communicate complex AI concepts clearly to engineering teams, business leaders, customers, security teams, and executive stakeholders.

- Make pragmatic build-versus-buy and model/platform decisions based on accuracy, scalability, latency, security, maintainability, and total cost.

- Drive architecture governance while enabling rapid experimentation and reusable accelerators.

- Coach cross-functional teams and promote disciplined engineering, documentation, evaluation, and responsible AI practices.

Preferred Industry Exposure

Industrial manufacturing, energy, utilities, engineering and construction, automotive, aerospace, asset-intensive operations, digital twins, IoT, predictive maintenance, or engineering modernization.

Success Measures

- AI solutions move from PoC to production with clear quality, safety, adoption, performance, and business-value metrics.

- CAD and engineering intelligence use cases reduce manual interpretation, improve retrieval and traceability, and accelerate engineering workflows.

- Agentic AI workflows demonstrate measurable task completion, groundedness, control, observability, and human accountability.

- Reusable architectures, components, and governance patterns shorten delivery cycles across programs.

- Platforms meet agreed standards for reliability, scalability, security, latency, and cost.

Ideal Candidate Profile The ideal candidate is an architect who can operate across the complete value chain: Business and engineering problem → data and drawings → model/LLM/VLM selection → Agentic AI and knowledge architecture → PoC → production deployment → MLOps/LLMOps → governance → measurable business value. The individual should be strategically strong, technically hands-on, and comfortable challenging assumptions while guiding teams toward practical, production-grade implementations.

Recognized as an expert within the organization/industry. Regarded as an authoritative source within a specialized discipline. Responsible for the development of software applications as a product or product line. Focus is on creation of the source code or application for commercial release. May include Applications Development positions as well. Provides insight on complex/large projects. Mentors junior staff but does not have direct supervisory responsibility.

📌 Principal Architect — Agentic AI (Hyderabad)
🏢 Cyient
📍 Hyderabad