AI Engineer-Voice AI, LLM Agents & Orchestration

Ignosis is a well-funded fintech infrastructure company building intelligent systems for the BFSI ecosystem. Backed by reputed investors, we focus on financial data intelligence, hyper-personalization, automation, and democratized credit access.

At Ignosis, we believe the next wave of fintech will be powered by AI systems that can think, speak, and act. We are building cutting-edge AI platforms that help financial institutions unlock deeper insights, automate complex workflows, and deliver smarter financial experiences.

Join us in building the future of AI-driven financial infrastructure.

Job Overview

We are building a production Voice AI Orchestrator — a system where LLM agents listen, reason, and respond in real time across voice and messaging channels.

As an AI Engineer, you will design and build LLM-powered agent systems, orchestration frameworks, and real-time voice pipelines. This is not a wrapper role — you will own the AI system architecture end-to-end, from agent design and evaluation to production reliability.

We are looking for engineers who have built and shipped AI systems in production, understand the challenges of non-deterministic LLM outputs, and enjoy solving complex real-world engineering problems.

Key Responsibilities

AI Agents & Orchestration

• Design and build multi-agent systems using frameworks like LangGraph and LangChain.
• Implement orchestration workflows with branching logic, retries, and parallel tool calls.
• Build intelligent pipelines with custom tools, memory layers, and dynamic routing.

Voice AI Systems

• Build real-time voice AI pipelines (STT → LLM → TTS) optimized for low latency.
• Integrate telephony systems via Twilio, WebRTC, and SIP.
• Engineer conversational voice experiences that perform reliably in production.

LLM Reliability & Evaluation

• Design LLM evaluation pipelines including regression testing and adversarial evaluation.
• Build guardrails, output validation, and fallback routing for reliable AI behavior.
• Implement observability and monitoring for LLM systems.

Retrieval & Knowledge Systems

• Build RAG pipelines for intelligent knowledge retrieval
• Implement hybrid search, re-ranking, and context compression strategies
• Work with vector databases and embedding systems

Backend & System Integration

• Integrate AI systems with Java / Spring Boot backend services
• Implement WebSocket and asynchronous APIs for real-time agent communication.
• Build scalable and reliable system architectures.

Experience

• 2+ years of software engineering experience
• Experience building AI or LLM-based applications
• Prior experience deploying LLM agents or AI pipelines into production environments

Agents Framework

Experience with one or more: LangChain/LangGraph/LlamaIndex/CrewAI/Haystack/Semantic Kernel

LLM Engineering

• OpenAI APIs / GPT models
• RAG (Retrieval-Augmented Generation)
• Prompt engineering and context design
• Function calling and structured outputs
• LLM evaluation and reliability techniques

Voice AI

STT / TTS systems/Whisper/Deepgram/ Azure Speech/Google Speech/Real-time voice pipelines/Telephony integration (Twilio, WebRTC, SIP)

Backend

Java · Spring Boot · WebSockets · Async / Multithreading · REST APIs

AI Infrastructure

• Vector Databases · Embeddings · Semantic Search
• Pinecone · Qdrant · Weaviate · Milvus · FAISS

Who We Are Looking For

• Engineers who have shipped AI systems to production
• Builders who think in systems — latency, evaluation, reliability
• People who understand that prompt engineering is engineering
• Engineers comfortable working in fast-moving, ambiguous environments
• Individuals who take full ownership from idea to production

Nice to Have

• Experience with open-source LLMs (Llama, Mistral, Gemma)
• Familiarity with LangSmith, Helicone, or LLM observability tools
• Experience with real-time voice or telephony AI
• Contributions to open-source AI tools

What's in it for you

• Build cutting-edge AI systems combining LLMs, voice AI, and agent orchestration
• Work on real-world fintech problems with top banks, NBFCs, and fintechs
• Be part of a fast-moving startup building next-generation AI infrastructure
• Competitive compensation and high ownership

To conclude, this role is ideal for engineers who want to build production-grade AI systems that think, speak, and act. At Ignosis, you will help shape the future of AI-powered fintech infrastructure.

To apply, send your resume to [Confidential Information]