Forward Deployed Engineer (FDE) - (India/Remote)

Voice AI & Prompt Engineering Craft Voice‑Optimized Prompt Flows: Design conversational flows that account for natural speech patterns (pauses, interruptions, intonation) optimized for voice-only interactions. Ensure prompts are clear for TTS pronunciation to avoid ambiguity. Continuous Refinement: Use LLM feedback loops and "self‑reflection" to score outputs, detect hallucinations, and improve prompts. Set up pipelines for A/B testing, prompt versioning, and performance QA tailored to financial use cases. Voice Integration: Collaborate with engineering teams to integrate prompts with speech recognition, intent extraction, LiveKit voice infrastructure, and telephony APIs. Ensure orchestration maintains real‑time responsiveness and low latency.

Craft Voice‑Optimized Prompt Flows: Design conversational flows that account for natural speech patterns (pauses, interruptions, intonation) optimized for voice-only interactions. Ensure prompts are clear for TTS pronunciation to avoid ambiguity.

Continuous Refinement: Use LLM feedback loops and "self‑reflection" to score outputs, detect hallucinations, and improve prompts. Set up pipelines for A/B testing, prompt versioning, and performance QA tailored to financial use cases.

Voice Integration: Collaborate with engineering teams to integrate prompts with speech recognition, intent extraction, LiveKit voice infrastructure, and telephony APIs. Ensure orchestration maintains real‑time responsiveness and low latency.

Agentic Architecture, LLMs & RAG Build Hierarchical Workflows: Develop finance AI agents that coordinate sub‑agents (e.g., a Research Agent, a Finance Agent, and an Editor Agent) for scalability and modularity. Model Optimization: Apply expertise in leading LLMs (Gemini, GPT series, Claude) to optimize our AI Finance Employee performance. Ensure low latency and high efficiency across all applications. Grounding & RAG: Integrate retrieval-augmented generation (RAG) with enterprise knowledge bases and financial APIs to prevent hallucinations and maintain tight context control around business domains. System Integration: Architect and integrate LLM systems with third-party tools, email interactions, and user chat interfaces. Develop complementary components like customizable OCR models.

Build Hierarchical Workflows: Develop finance AI agents that coordinate sub‑agents (e.g., a Research Agent, a Finance Agent, and an Editor Agent) for scalability and modularity.

Model Optimization: Apply expertise in leading LLMs (Gemini, GPT series, Claude) to optimize our AI Finance Employee performance. Ensure low latency and high efficiency across all applications.

Grounding & RAG: Integrate retrieval-augmented generation (RAG) with enterprise knowledge bases and financial APIs to prevent hallucinations and maintain tight context control around business domains.

System Integration: Architect and integrate LLM systems with third-party tools, email interactions, and user chat interfaces. Develop complementary components like customizable OCR models.

Data Analytics & Workflow Automation Process Improvement: Analyze business processes, data quality, and operational bottlenecks to identify improvement opportunities and present actionable recommendations. Automation & Reporting: Automate workflows, data transformations, and reporting using Python and SQL. Build and maintain dashboards and data monitoring frameworks. Collaboration & Documentation: Work closely with product, engineering, and operational teams to define requirements. Document workflow logic, scripts, and solution designs clearly.

Process Improvement: Analyze business processes, data quality, and operational bottlenecks to identify improvement opportunities and present actionable recommendations.

Automation & Reporting: Automate workflows, data transformations, and reporting using Python and SQL. Build and maintain dashboards and data monitoring frameworks.

Collaboration & Documentation: Work closely with product, engineering, and operational teams to define requirements. Document workflow logic, scripts, and solution designs clearly.