Our client is a well-funded AI-based EdTech partnering closely with multiple government entities to deliver learning solutions impacting students at a population scale.

● Develop and implement state-of-the-art NLP, Computer Vision, Speech, GenAI

techniques.

● Design, build, fine-tune, and evaluate LLMs, VLMs, and SLMs for conversational,

multimodal, and educational use cases.

● Conduct research and experiments in deep learning, machine learning, Generative

AI, and foundation models to support the development of scalable AI techniques.

● Lead the instruction fine-tuning of Small Language Models (SLMs) (utilizing

LoRA/PEFT) on proprietary datasets to achieve state-of-the-art performance.

● Architect deep temporal models (LSTMs/Transformers/DMSW) that ingest

multimodal behavioral data to build predictive models (e.g. predict student dropout

risks and learning gaps weeks in advance).

● Design, implement, and optimize robust Multi-Agent Systems where specialized

agents (Tutors, Retrievers, Critics) collaborate to solve complex pedagogical tasks.

● Architect end-to-end AI pipelines capable of handling massive throughput (10M+

daily active users) with sub-second latency constraints. Optimize the high-stakes

trade-offs between model complexity (accuracy), inference latency, and unit

economics.

● Mentor junior AI Engineers and work with data scientists, fostering a culture of

technical excellence.

● Experience: 5+ years in Software Engineering, with at least 5+ years deeply focused

on AI/ML. A proven track record of designing and deploying AI models to production

at scale (millions of users).

● Education: MS or PhD in Computer Science, AI or a related field with experience in

machine learning and data analysis.

● Generative AI Stack: Deep expertise in the modern LLM ecosystem: Transformers

(Hugging Face), Orchestration frameworks (LangChain, LangGraph, LlamaIndex),

Vector Databases (Milvus, Pinecone, Weaviate), and Inference Optimization (vLLM,

TGI).

● Assessment Technology: Experience with Optical Character Recognition and

psychometric models (Item Response Theory, DKT) is a significant differentiator.

● Hands-on experience in ML model development, fine-tuning, and deployment,

including productionizing AI/GenAI systems.

● Deep Learning Mastery: Expert-level proficiency in PyTorch or TensorFlow. Strong

theoretical and practical understanding of sequence models (RNNs, LSTMs,

Transformers) and their application in NLP and Time-Series forecasting.

● Strong programming skills in Python, node.js or other relevant programming

languages.

● Familiarity with machine learning libraries, such as TensorFlow, Keras, PyTorch,

scikit-learn, Conversational frameworks like Rasa, DialogFlow, Lex, etc.

● Data & Cloud Engineering: Experience with big-data processing and cloud-native AI

infrastructure (AWS Bedrock, SageMaker, or GCP Vertex AI) is a huge plus.

● Strategic Competencies: A "Systems Thinking" mindset. Exceptional ability to

communicate complex AI concepts to non-technical stakeholders (government

officials, product leaders). A demonstrated passion for mentorship and team building.

An opportunity to create a impact at nation-level, working closely with multiple government bodies.

This is a high-ownership, hands-on leadership role at the intersection of deep-tech architecture, public-scale delivery, and mission-critical reliability.