About Us Gramian Consultancy is a boutique consultancy specializing in IT professional services and engineering talent solutions. With a strong background in software engineering and leadership, we help companies build high-performing teams by matching them with professionals who truly fit their needs. Role overview We are looking for technically skilled professionals to join a short-term project as Technical Annotators, contributing to the creation and validation of high-quality workflow data used to train advanced AI systems. In this role, you will produce detailed, step-by-step technical workflows for developer-adjacent tasks such as file operations, scripting, API usage, terminal commands, and browser-based automation. The work requires strong attention to detail and the ability to translate natural-language instructions into precise executable steps. This role is ideal for candidates with experience in software development, technical support, scripting, or Linux environments. Commitments Required: 6-8 hours per day with an overlap of 4 hours with PST. Employment type: Contractor assignment (no medical/paid leave) Duration of contract: 5 weeks Location: India, Pakistan, Nigeria, Kenya, Egypt, Ghana, Bangladesh, Turkey, Mexico, Brazil Interview: take home assessment (60min) Your Daily Responsibilities - Create step-by-step technical workflows based on natural language instructions - Produce structured task trajectories for developer-style activities - Execute and document workflows in Linux desktop environments - Validate trajectories for correctness, completeness, and reproducibility - Use Python or Bash scripting to simulate or verify task execution - Work with APIs, terminal commands, files, and browser-based tools - Review annotations to ensure consistency and high quality - Follow detailed annotation guidelines and quality standards - Collaborate with internal reviewers to improve task accuracy

• Serve as the lead AI Engineer as part of Data Science & Advanced Analytics. • Act as a liaison between management, technical teams, and stakeholders, communicating decisions and technical information. • Architect, develop, and maintain intelligent AI agents using Copilot Studio and Azure AI (OpenAI, Bot Service, Machine Learning). • Optimize AI agents for performance, scalability, and reliability. • Monitor and evaluate AI solutions, ensuring alignment with business goals. • Build and manage Azure Functions, Logic Apps, and automation workflows. • Orchestrate large language models (LLMs) for enterprise use cases (conversational AI, summarization, decision support). • Document and oversee project tasks, maintain project files, and support client meetings. • Develop and adhere to project timelines, identify process improvements, and implement quality initiatives. • Prepare technical documentation, reports, and deliverables using Microsoft Word and Excel. • Conduct research on AI trends, best practices, and emerging technologies. • Participate in training and educational presentations as required. • Monitor and optimize AI system performance, scalability, and reliability. • Ensure compliance with security, privacy, and governance standards. • Always represent HSAG in a professional manner.

• Apply your knowledge of AI systems and software engineering to develop solutions that directly address and resolve business problems. • Partner with professionals from Data Science and Data Engineering to address complex technical challenges, ensuring that the latest and most effective Data & AI techniques are being utilized. • Take ownership of implementing and optimizing applied AI components, ensuring they meet project needs with high complexity and scale. • Navigate and manipulate generative AI models, including (but not limited to) LLMs, to create prompts and solutions tailored to specific use cases. • Develop and incorporate AI solutions while adhering to industry best practices, including moderation, security, monitoring, and compliance standards. • Understand and properly apply Responsible AI concepts in all the stages of the solution. • Lead the charge in designing, measuring, and evaluating AI model outputs, developing standard and custom metrics to ensure alignment with business objectives. • Translate AI research and PoCs into production-ready features, delivering robust and scalable AI components that integrate seamlessly with larger systems. • Drive the selection and application of appropriate evaluation metrics, ensuring that AI solutions are robust, unbiased, and meet all necessary performance standards.

• Architecting and Implementing Agentic Workflows • Design Autonomous Loops: Build stateful control loops (Perception -> Reasoning -> Action) where agents can interpret clinical data, plan necessary checks, and generate risk assessments. • Conversational Agentic workflow: Chatbot style user experience powered by LLM • GCP Integration: Leverage Vertex AI for model serving and LLM models for reasoning, optimizing for long-context windows to process complex medical history. • RAG & Grounding: Implement "Enterprise Grounding" using Vertex AI Vector Search to ensure all agent outputs are strictly cited against official medical literature (e.g., medical research and guidelines). • Data & Tool Integration • Structured Data Access: Build robust Python tools (Function Calling) that allow agents to securely query BigQuery or other datastores for patient demographics and symptoms. • Security & Compliance: Ensure all AI operations comply with HIPAA/GDPR standards, implementing strict PII masking and data governance within the Google Cloud environment. • System Reliability: Deploy agents on Vertex AI, Cloud Run, or GKE, ensuring low-latency inference and high availability for clinical users. • Agent monitoring: implement effective product environment AI Agent execution monitoring and alert system to ensure SLA, safety and security compliance • Leadership & Strategy • Mentorship: Guide a team of engineers in MLOps best practices (CI/CD for models, evaluation pipelines). • Evaluation: Implement "LLM-as-a-Judge" frameworks to automatically test agent accuracy against "Golden Datasets" of clinical cases.