Ascendion - Artificial Intelligence Architect - LLM/Prompt Engineering

Job Description : Role : AI Architect. Location : PAN India (Remote/Work From Home). Enterprise-Scale AI Solution Design : - Designs full-stack conversational AI solutions using Google CCAI, Vertex AI, and custom LLM pipelines. - Leads architecture for virtual agents (text/voice), agent assist, and AI-driven analytics Generative AI (Vertex AI, PaLM, GPT-4). - Designs and governs GenAI applications : summarization, Q&A, document intelligence. - Manages model selection, prompt architecture, grounding strategies, and safety guardrails. - Conversational Architecture (Dialogflow CX Voice). - Oversees state machine design, omnichannel integration, and multilingual bot scaling Architects fallback logic, entity modeling, webhook orchestration, and call flow transitions. Data Engineering & Pipeline Strategy : - Designs ingestion and transformation flows using BigQuery, Dataflow, Pub/Sub, Cloud Storage. - Aligns data pipelines to support feature engineering and real-time inference. MLOps & Continuous Deployment : - Implements scalable training/inference workflows using Vertex Pipelines, TensorFlow Extended, or Kubeflow, Versioning, CI/CD, reproducibility, and rollback capabilities built into the ML lifecycle. Knowledge Base & Embedding Systems : - Integrates unstructured data pipelines using LangChain, LlamaIndex, Pinecone, Vector Search. Google Agent Assist & Speech Systems : - Configures Agent Assist triggers, STT tuning (noise handling, custom lexicons), and Smart Reply modules. - Architects seamless escalation logic and real-time transcription pipelines. Voice AI Design : - Leads telephony integration (SIP, PSTN), IVR workflows, and low-latency voice bot optimization. - Balances TTS/STT integration with call handling logic and human-agent transfer workflows. Testing & Validation Strategy : - Defines testing standards for LLM responses, STT accuracy, webhook logic, and CI automation. - Leads model evaluation, prompt regression testing, and safety testing across pipelines. AI Governance & Compliance : - Implements data privacy standards (e.g. , HIPAA, GDPR), PII handling, and audit logging. - Establishes ethical AI use guidelines, fallback safety protocols, and model monitoring. Executive-Level Reporting : - Defines KPIs for AI success : containment rate, CSAT lift, time-to-resolution, and NPS. - Oversees dashboards (Looker, Tableau) and insight generation for product, ops, and leadership. Cross-Functional Collaboration : - Leads teams of developers, analysts, QA, and business leads through Agile execution. - Conducts architectural reviews, sprint planning, and stakeholder demos. Tools & Platforms : - Languages : Python, SQL, Node.js. - Cloud : GCP (Vertex AI, BigQuery, Dialogflow CX, Cloud Functions, Cloud Run, IAM). - AI Tools : LangChain, Pinecone, LlamaIndex, OpenAI APIs. - Testing : Postman, Selenium, PyTest, JMeter, CI/CD (Cloud Build/GitHub Actions). - Visualization : Looker, Data Studio, Tableau. - Project Management : Jira, Confluence, Figma (for conversational UX review). (ref:hirist.tech)

Location: in, IN

Posted Date: 5/24/2025