Vtricks Technologies
Website:
vtricks.in
Job details:
Senior Software Developer – LLM Training & AI Solutions
Location: Bengaluru (1 Month Work From Office) → Remote (Next 5 Months)
Experience: 8–12+ Years
Budget: 2 LPM
Role Overview
We are looking for a Senior Software Developer with deep expertise in Large Language Models (LLMs) to lead the end-to-end development of domain-specific AI solutions.
This is a hands-on technical leadership role focused on training, fine-tuning, evaluating, and deploying LLMs tailored to enterprise use cases across industries such as education (SIS), manufacturing, and industrial systems.
You will build scalable AI systems powering web apps, mobile assistants, and voice agents, ensuring models are aligned with domain-specific workflows, compliance, and business needs.
Key Responsibilities
LLM Training & Fine-Tuning
- Design and execute LLM training pipelines including:
- Supervised Fine-Tuning (SFT)
- RLHF / DPO
- Continual pre-training
- Implement PEFT techniques (LoRA, QLoRA, Adapters)
- Build multi-tenant model architectures with versioning and isolation
Data Pipelines & Infrastructure
- Build pipelines for:
- Data ingestion, cleaning, labeling, and transformation
- Ensure data quality, validation, and bias control
- Set up scalable training infra on:
- AWS SageMaker / Azure ML / GCP Vertex AI
- Enable distributed training and experiment tracking (MLflow, W&B)
Model Evaluation & QA
- Develop evaluation frameworks using:
- RAGAS, DeepEval, TruLens
- Implement:
- Hallucination detection
- Bias & fairness checks
- Regression and performance testing
- Drive A/B testing and production validation
Application Integration
- Build and integrate LLMs into:
- Web apps (APIs, copilots)
- Mobile apps (optimized inference)
- Voice agents (STT → LLM → TTS pipelines)
- Work with:
- Whisper, Deepgram, ElevenLabs, Azure TTS
MLOps & Multi-Tenant Systems
- Build CI/CD pipelines for ML models
- Implement:
- Model monitoring & alerting
- Automated retraining pipelines
- Manage tenant-specific models and data isolation
Required Skills
Core Technical Skills
- Python (Expert level)
- Strong experience in LLM training & fine-tuning (2–3+ years)
- Hands-on with:
- Hugging Face Transformers, PyTorch, DeepSpeed
- LangChain / LlamaIndex
- Experience with:
- RAG pipelines & vector databases (Pinecone, Weaviate, pgvector, etc.)
AI/ML & Data
- Strong understanding of:
- NLP, Deep Learning, Model Optimization
- Experience in:
- Data pipelines (Airflow, Spark, dbt)
- Data labeling and curation
Cloud & Infrastructure
- Hands-on experience with:
- AWS / Azure / GCP
- Knowledge of:
- Docker, Kubernetes, GPU environments
MLOps
- Experience with:
- MLflow, Weights & Biases
- CI/CD for ML pipelines
- Model monitoring & versioning
Integration & APIs
- Experience building APIs using:
- FastAPI / Flask
- Understanding of:
- Web/mobile integration (React / Flutter – basic exposure sufficient)
Voice AI (Preferred)
- Experience with:
- STT, TTS, real-time streaming pipelines
- Voice agent frameworks
Preferred Qualifications
- Experience with:
- Open-source LLMs (LLaMA, Mistral, Falcon, etc.)
- Model optimization (quantization, pruning)
- Exposure to:
- EdTech (SIS) or Manufacturing domains
- Experience with:
- Agentic AI frameworks (AutoGen, CrewAI, LangGraph)
- Knowledge of:
- Compliance frameworks (FERPA, ISO standards)
Key Competencies
- Strong problem-solving and system design skills
- Ability to own end-to-end AI lifecycle
- Strong communication and stakeholder collaboration
- Experience working in fast-paced, product-driven environments
Click on Apply to know more.