Job Description
About The Company: QualityKiosk Technologies is one of the world's largest independent Quality Engineering (QE) providers and digital transformation enablers, helping companies build and manage applications for optimal performance and user experience. It is one of the worlds largest independent quality assurance providers
Website: www.qualitykiosk.com
Please find below job details fyr:
Job location: Machine Learning Engineer - Bangalore
JD # Machine Learning Engineer
We are looking for a Machine Learning Engineer with deep, hands-on experience training and fine-tuning open-source LLMs, and strong competence in working with foundation models as programmable systems.
This role is responsible for building state-of-the-art, domain-optimized open-source models tailored for our AI Evaluation Platform and AI Test Execution Platform.
You will not only improve model weights through training and fine-tuning, but also design structured LLM workflows and evaluation-driven optimization loops that improve real-world model behavior.
This is not an API-integration role.
This is a hands-on role requiring real experience in model training, benchmarking, structured prompting, and systematic LLM optimization.
Key Responsibilities
- Train, fine-tune, and benchmark open-source LLMs and ML models
- Design models optimized for:
- AI evaluation and scoring
- Classification and grading tasks
- Test generation and adversarial analysis
- Own the full model lifecycle:
- Dataset preparation and curation
- Training and fine-tuning
- Validation and benchmarking
- Error analysis and iterative improvement
- Work with foundation models as programmable systems, not just black-box APIs
- Design and optimize structured LLM workflows, including:
- Multi-step reasoning pipelines
- Prompt optimization loops
- Evaluation-driven refinement approaches
- Build evaluation and validation pipelines assessing:
- Performance
- Robustness
- Reliability
- Bias and safety metrics
- Collaborate closely with:
- AI QA teams to align evaluation criteria
- Product engineers for system integration
- Platform engineers for training and deployment constraints
- Optimize models for:
- Practical deployability
- Inference cost
- Latency
- Real-world robustness
Must-Have Skills & Experience
- 510+ years in applied ML or ML engineering
- Proven experience training and fine-tuning open-source LLMs
- Hands-on experience with models such as:
- LLaMA variants
- Mistral / Mixtral
- Falcon
- Other open-source transformer models
- Strong proficiency in:
- Python
- PyTorch (required)
- Hugging Face ecosystem
- Experience with:
- Supervised fine-tuning (SFT)
- LoRA / QLoRA / adapter-based approaches
- Model benchmarking and structured evaluation
- Experience working with foundation models as programmable systems
- Familiarity with structured LLM optimization approaches (e.g., DSPy-style workflows, prompt compilation, or evaluation-driven optimization)
- Strong understanding of:
- Transformer architecture fundamentals
- Training vs inference trade-offs
- GPU constraints and optimization
- How prompt design, training, and evaluation metrics interact
- Experience writing production-grade ML code
kindly revert with your updated CV and below details at email ID - clive.lobo@qualitykiosk.com
Full Name:
Total Exp:
Relevant exp in Years - Machine Learning/ Python/ LLMs...??-
Current Company:
CTC:
ECTC:
Notice Period:
Reason for job change:
Current Residential location:
Ready to work from office?
Preferred job location: Bangalore ,
Any offers in Hand? (Please share details CTC, LWD, DOJ):
Thanks & Regards,
Clive Lobo
QualityKiosk Technologies
clive.lobo@qualitykiosk.com