Ogha Technologies · 9 hours ago
Senior AI/ML Engineer – LLMs & Fine‑Tuning
San Jose, CA
Full-time
Onsite
Mid, Senior Level
3+ years expOgha Technologies is seeking a senior AI/ML engineer with deep experience in large language models (LLMs). The role involves owning the end-to-end development of LLM-powered features, including data pipelines, training, evaluation, and integration with products.
AnalyticsArtificial Intelligence (AI)Big DataConsultingData IntegrationInformation TechnologyNatural Language Processing
Responsibilities
Design, train and fine‑tune LLMs (e.g., GPT‑class, Llama, Mistral, Qwen) using techniques such as LoRA/QLoRA, instruction tuning, and domain adaptation on proprietary datasets
Build data pipelines for collecting, cleaning, labeling, and augmenting text datasets for supervised fine‑tuning, preference modeling, and evaluation
Develop and maintain scalable training and inference pipelines using frameworks such as PyTorch, TensorFlow/JAX, Hugging Face Transformers, and associated tooling
Implement and optimize RAG (Retrieval‑Augmented Generation) systems leveraging vector databases (e.g., FAISS, Pinecone, Weaviate, Qdrant) and document stores
Design prompt strategies, tools, and agents to improve reliability, controllability, and latency of LLM‑based applications
Define and implement evaluation frameworks (automatic metrics, human evals, red‑team tests) for quality, safety, and robustness of model outputs
Optimize models for inference via quantization, distillation, pruning, and efficient serving (GPU/CPU, batch inference, caching)
Collaborate with product, engineering, and domain experts to translate business problems into LLM‑based solutions and iterate quickly on prototypes
Deploy, monitor, and maintain LLM services in production using MLOps best practices (CI/CD, experiment tracking, model/version management, A/B testing)
Proactively track state‑of‑the‑art research in LLMs, multimodal models, and alignment, and bring relevant advances into our stack
Qualification
Required
Bachelor's or Master's degree in Computer Science, Machine Learning, Mathematics, or related technical field (or equivalent practical experience)
Strong Python programming skills and software engineering fundamentals (testing, code review, modular design, performance profiling)
3+ years of hands‑on experience building ML systems end‑to‑end, with at least 1–2 years focused specifically on NLP/LLMs
Proven experience fine‑tuning LLMs (parameter‑efficient methods and/or full‑parameter) on real‑world datasets and shipping them into production
Deep understanding of modern deep learning and transformer architectures: attention mechanisms, positional encodings, tokenization, optimization (AdamW, schedulers), and regularization
Experience with PyTorch (preferred) or TensorFlow/JAX, plus Hugging Face ecosystem (Transformers, Datasets, Accelerate, PEFT, TRL, etc.)
Experience building and operating training and inference workloads on cloud platforms (AWS, GCP, Azure) and GPUs (CUDA, distributed training with DDP/DeepSpeed/Lightning, etc.)
Strong skills in designing and querying data stores for RAG (SQL/NoSQL, vector databases) and integrating them with LLMs
Familiarity with deploying models as APIs/microservices using frameworks such as FastAPI, Flask, or similar
Solid understanding of model evaluation, observability, and monitoring (quality, drift, bias, safety)
Preferred
Experience implementing RAG systems at scale (indexing, retrieval optimization, hybrid search, metadata‑aware ranking)
Experience with RLHF, DPO, or other alignment techniques and preference‑based training
Experience with multi‑modal models (e.g., vision‑language, speech‑language) or tool‑using/agentic architectures
Exposure to security, compliance, and privacy aspects of training and serving LLMs on sensitive data
Contributions to open‑source ML/LLM projects, publications, or notable public demos
Experience integrating with major LLM APIs and managed services (OpenAI, Anthropic, Google, Azure, etc.) as well as self‑hosted models
Ability to work closely with non‑ML engineers, PMs, and stakeholders, explaining complex concepts clearly and pragmatically
Product mindset with focus on impact, reliability, and user experience rather than just model metrics
Self‑driven, comfortable with ambiguity, and able to own projects from idea to production
Company
Ogha Technologies
Ogha Technologies is an IT company that provides big data, data integration, advanced analytics, and IT strategy consulting services.
11-50 employees
Funding
Current Stage
Early StageLeadership Team
Bhanu Murthy Nallagonda
Co-Founder and Director
Company data provided by crunchbase