Ahana Sarkar

Product Management: Deliver AI solutions with in-depth knowledge of the product development lifecycle, Roadmap Planning, Feature Prioritization, Stakeholder Management, Product Metrics to optimization techniques for Bot Operations, with an Agile and Kanban approach. Conversational Experience Design: Adapt Bot’s understanding and communication to the specific user needs, business context, and channels (e.g. voice vs. chat, web). Create sample dialogues, conversational flow diagrams, and prototypes to effectively communicate voice and text interactions and design ideas.

Intent recognition Design: Generate and train intents to Amelia for better user intention recognition, Named Entity Recognition (NER), quality tests through sentiment analysis, regression test, using LangChain and other alternative LLM frameworks to increase intent accuracy. AI Training: Train data using various ML models for Deep Learning projects using CNN, GAN, Transformers, Encoder and decoder algorithms to test Amelia’s classification engines and natural language understanding (both corpus level and domain-specific level). Design NLU and Prompt Engineering: Implementing natural language understanding (NLU) and ML algorithms in the Amelia platform to identify and refine user personas and problem statements, Few shot test, intent score mapping, generating data, chain-of-thoughts and more. Post Live QA Review: Quantitative and qualitative insights from UAT testing and set up subsequent design sprints for continuous monitoring of containment rate, new intent mapping, training and handover to diverse stakeholders.

User Research: Create unique Bot Persona and human like conversational elements to suit the user’s temperament and conversation style after analysing previous call records, use-case planning and test case generation. Back-end Process: Build the bot with various Indian linguistic variations of expected user Intents to optimize bot intelligence, call tagging, API Integration and generating YAML codes, Wireframing and Prototyping and Project documentation. Designing the framework: Create and define concept and execution, articulating end-to-end experience, and help drive that vision into solid design deliverable’s. Interface designing for the Voice agent: Led system personality efforts to anchor voice bot consistency, voice, tone, cognitive load, pragmatics, prosody, speech and language technologies (e.g.ASR, NLU, TTS). Post-live review analysis: Optimizing intents through A/B testing, usability, and QA testing to make design recommendations, analyze performance metrics, like NLU accuracy, Unique leads, transferred leads, total calls dialed, KPI and OKR’s to constantly match the revenue goals. Result Efficiency: In a year, my bots have answered 42,000+ customer calls, generated 20,000+ unique leads, increased 20% customer conversion and reduced 80% calling efforts of the human agents.