AI Engineer

Overview

Location: Bengaluru
Type: Full-time
Compensation: ₹15 - ₹25 LPA (experience & skills-based; performance-linked growth and incentives)

About Samora
Samora builds enterprise-grade voice agents that sound natural, act intelligently, and scale across thousands of conversations. We focus on real-world reliability — low latency, high accuracy, personalization, self learning and human-in-the-loop workflows. We’re backed by leading organizations (Microsoft, OpenAI, Meta, AWS, Deloitte, UNICEF, and more) and built by leaders from Stanford, Microsoft, Uber, and IBM Research.

Why this role
You will work at the frontier of applied voice AI — evaluating the newest research, testing models, running experiments, and building capabilities that provide voice agents with the human touch. You will help us improve accuracy, reduce latency and cost, optimize tool-calling, and push the boundaries of conversational AI.
You will spend time both experimenting with frontier tech and shipping production-grade improvements.

What you’ll do

1. Build, fine-tune, evaluate, and deploy LLM/ASR/TTS-based systems for voice agents
2. Research and test the latest AI techniques, models, and tooling
3. Improve latency, accuracy, quality, and cost of inference
4. Enhance tool-calling, memory, and knowledge-retrieval workflows
5. Integrate and benchmark model providers, embeddings, and RAG pipelines
6. Work closely with founders to prioritize improvements that move key metrics
7. Build frameworks for automated testing, evaluation, and iteration of voice systems

You’ll be great if you

1. Have strong fundamentals in ML/LLMs, deep learning, NLP, and/or speech systems
2. Understand or are eager to learn about real-time speech pipelines
3. Are comfortable reading research papers and rapidly prototyping
4. Can balance rapid experimentation with production execution
5. Enjoy owning problems end-to-end and thinking on a long-term tech horizon
6. Are self-driven with curiosity about speech, LLMs, and automation

Tech stack exposure (plus to have)
Experience in some of the following is useful — not required to know all:

1. Python, PyTorch, Transformers ecosystem
2. LLM frameworks (HuggingFace, LangChain/LangGraph, OpenAI APIs, vLLM, Ollama)
3. Speech tech: ASR/TTS models (Whisper, Deepgram, PlayHT, ElevenLabs, NeMo, Coqui, etc.)
4. Vector DBs & RAG (Pinecone, Weaviate, Chroma)
5. Evaluation frameworks & benchmarking
6. Realtime audio pipelines (WebRTC/LiveKit/Twilio/Plivo) is a plus
7. Experience with GPU inference, model quantization, distillation is a bonus
8. Previous voice AI or conversational AI projects are a strong advantage

Who can apply

1. Strong engineering & problem-solving fundamentals
2. Exceptional college beginners with projects in AI/LLMs
3. Passion for speech, LLMs, or real-time AI systems

Growth & Path

1. Work with founders with enterprise AI experience on frontier voice AI problems
2. Exposure to production workflows at scale
3. High-ownership role with path to lead AI systems