AI Engineer - Foundation Models & Sovereign AI Systems

We are looking for AI Engineers with 1-2 years of strong hands-on experience to work on foundation model systems across language, multimodal, and speech AI. The role is focused on building sovereign AI capabilities for enterprise and client use cases, where control over data, models, infrastructure, deployment, and governance is critical.

This is a builder role for candidates who understand modern AI systems conceptually and technically, and can convert fast-moving model developments into reliable working solutions.

●     Build and evaluate solutions using open-source and enterprise-grade LLMs, multimodal models, speech models, embedding models, and retrieval systems.

●     Work on model adaptation, fine-tuning, LoRA/QLoRA, prompt optimization, RAG pipelines, agentic workflows, and evaluation frameworks.

●     Compare models across quality, latency, cost, accuracy, safety, deployment constraints, and client-specific requirements.

●     Develop AI applications that can run in controlled environments, including private cloud, on-premise, or client-owned infrastructure.

●     Create and maintain clean experimentation workflows for datasets, model benchmarks, inference behavior, and failure analysis.

●     Collaborate with engineering and product teams to convert AI prototypes into deployable systems.

●     Track relevant model releases, architecture improvements, benchmarks, and tooling changes in the AI ecosystem.

●     1-2 years of practical experience in AI/ML, NLP, LLMs, speech AI, computer vision, or multimodal AI.

●     Strong Python programming skills and comfort building production-oriented AI workflows.

●     Hands-on experience with PyTorch, Hugging Face, vector databases, embedding models, model APIs, or inference frameworks.

●     Strong conceptual understanding of transformers, tokenization, embeddings, attention, fine-tuning, RAG, context windows, hallucination, and model evaluation.

●     Ability to debug model behavior, analyze outputs, improve prompts or pipelines, and identify practical limitations.

●     Familiarity with APIs, Git, Docker, cloud platforms, GPU environments, and basic MLOps practices.

●     Clear communication skills and the ability to explain technical tradeoffs to both engineering and business teams.

●     Experience with ASR, TTS, voice agents, diarization, or speech-to-speech systems.

●     Experience with multimodal models for image, video, document, or visual reasoning use cases.

●     Exposure to vLLM, TensorRT-LLM, ONNX, quantization, Triton, CUDA, or inference optimization.

●     Experience deploying AI systems in private, secure, or regulated environments.

●     Strong GitHub, open-source work, serious personal projects, or prior AI product experience.

The ideal candidate is technically curious, execution-focused, and comfortable working in a fast-changing AI landscape. They should not simply use AI tools; they should understand how modern models work, where they fail, how to evaluate them, and how to integrate them into real systems.

We are looking for people who can think deeply, build quickly, question assumptions, and take ownership of complex AI problems.

Candidates should be able to demonstrate at least one meaningful AI project involving model evaluation, fine-tuning, RAG, speech AI, multimodal AI, or deployment. They should be prepared to explain their technical choices, tradeoffs, failure cases, and how they would improve the system.