AI Research Scientist

About US Shunya Labs is building the Voice AI Infrastructure Layer for Enterprises powering speech intelligence, conversational agents, and domain-specific voice applications across industries. Born from deep work in mental-health AI and built for global enterprise scale, our stack combines state-of-the-art ASR/TTS models with an open-weights philosophy , driving accuracy, privacy, and scalability.

About the Role We’re looking for an

AI Research Scientist

who combines deep technical understanding of AI systems with the curiosity to explore how emerging models can reshape real-world industries. This role sits at the intersection of

research, strategy, and applied innovation

identifying automation opportunities, developing proofs-of-concept, and translating cutting-edge AI into scalable products. You’ll work closely with leadership, engineering, and clients to define the company’s

AI product roadmap , continuously evaluate the evolving AI landscape, and design systems that deliver measurable operational impact.

Key Responsibilities -

Industry Research & Opportunity Discovery Study operational workflows across diverse industries (insurance, healthcare, logistics, financial services, manufacturing, etc.). Identify

high-impact automation and augmentation opportunities

using AI and LLM-based systems. Conduct

process diagnostics

— mapping inefficiencies, manual dependencies, and automation gaps. Collaborate with domain experts to design

AI-driven transformation blueprints

for clients and internal use cases.

AI Research & Model Development Continuously track

latest AI research , model architectures, and open-source innovations (LLMs, vision models, multimodal systems). Conduct deep technical analysis of foundation models — transformer internals, embeddings, retrieval architectures, diffusion, etc. Build and fine-tune custom AI models for tasks such as language understanding, summarization, transcription, classification, and reasoning. Experiment with

emerging architectures

(Mixture of Experts, RAG, GraphRAG, multi-agent systems) to enhance accuracy and adaptability. Develop evaluation frameworks for model performance, fairness, and interpretability.

Productization & Innovation Work with engineering teams to translate research prototypes into

market-ready products . Define technical strategy and architecture for AI-driven product lines. Contribute to internal IP creation — model enhancements, pre-training pipelines, or domain-specific datasets. Lead feasibility assessments for model-hosting infrastructure, inference optimization, and deployment scalability.

Thought Leadership & Continuous Learning Stay ahead of global AI trends by following leading research labs, open-source communities, and academic breakthroughs. Produce internal whitepapers, presentations, and tech briefs summarizing emerging technologies. Represent the organization in conferences, panels, and AI research forums. Mentor junior researchers and engineers in AI methodologies and experimentation.

Strategic Collaboration & Solutioning Partner with business and product leaders to align research outcomes with commercial goals. Participate in

client discussions , understanding pain points and shaping AI-powered solutions. Provide strategic inputs on go-to-market initiatives for AI offerings.

Required Skills - Strong foundation in

AI/ML algorithms ,

deep learning architectures , and

transformer-based models . Experience with

Python ,

PyTorch ,

TensorFlow ,

Hugging Face , and

LangChain

ecosystems. Hands-on experience with

model training, fine-tuning, and evaluation . Ability to read and interpret

AI research papers , reproduce experiments, and assess feasibility. Understanding of

LLM internals

(tokenization, attention, context windows, embeddings, quantization). Strong analytical skills to connect

technical innovation with business value . Excellent written and verbal communication for presenting research and influencing stakeholders.

Nice to Have - Experience building

domain-specific AI models

(e.g., financial document analysis, claims automation, customer service bots). Exposure to

retrieval-augmented generation (RAG) ,

graph-based reasoning , or

multi-agent orchestration . Background in

data curation ,

synthetic data generation , or

evaluation pipelines . Familiarity with

AWS Sagemaker ,

Vertex AI , or

custom training environments . Published research or contributions to open-source AI projects.

Soft Skills - Strong curiosity and comfort with

ambiguous, open-ended problems . Strategic mindset — can balance research creativity with product pragmatism. Ability to distill complex technical ideas into business-friendly narratives. Collaborative and entrepreneurial spirit; thrives in fast-moving innovation environments.