PHD researcher - AI and Data Systems

Company Description

Echos empowers enterprises to leverage cutting-edge technologies to deliver scalable intelligence and transformative results. Backed by a combination of robust foundations, advanced AI acceleration tools, and strong industry partnerships, Echos provides solutions that ensure immediate returns on investment while building capabilities for the future. The company specializes in converting deep tech innovations into impactful, measurable outcomes that drive business success.

Role Description

You will help design and develop next-generation enterprise reasoning systems, including:

• Agentic AI architectures: multi-step thinking, planning, self-critique, simulation loops.
• Cognitive models: memory, belief revision, self-evaluation, knowledge graphs, world models.
• Enterprise-grade reasoning for banking, insurance, financial crime, fraud, and decision automation.
• Hybrid symbolic + neural systems, tool-augmented LLMs, and grounded action models.
• Contextual long-horizon reasoning across millions of data points.
• Evaluation frameworks for truthfulness, coherence, reasoning quality, and factual recall.
• Novel training pipelines for agents capable of judgment, auditability, and introspection.

You will be expected to invent, publish, translate theory into production, and push what enterprise AI is capable of.

Qualifications

Required Background

• Strong research experience in Machine Learning, NLP, Deep Learning, or Cognitive AI
• Understanding of LLMs, Transformers, memory architectures, or multi-agent systems
• Ability to reason about limitations of current AI systems and propose new architectures
• Experience with at least one: PyTorch, JAX, TensorFlow
• Strong grasp of theoretical concepts such as:
• Reasoning, planning, decomposition
• Causality, uncertainty, Bayesian models
• Long-context models and retrieval systems
• Neural-symbolic methods
• Experience with research papers (author or co-author is a plus)

Plus:

• Experience building agentic or tool-using models
• Experience with financial, insurance, or enterprise workflows
• Prior publications in NeurIPS, ICML, ICLR, ACL, etc.
• Complex system design or distributed training experience

Role and Responsibilities:

• You help design a reasoning loop better than chain-of-thought
• You create an agent architecture that self-corrects and uses tools intelligently
• You contribute to papers, patents, and new model designs
• You help Echos build the world’s most trusted enterprise thinking engine

What we Offer:

• Work on some of the most challenging problems in AI reasoning
• The chance to define and shape enterprise intelligence systems globally
• A fast-moving, research-first environment
• Competitive salary + Other benefits
• Direct impact on next-generation industry AI systems