Fermin Agudo
Data Scientist
5+ projects
Python
Active 2 days ago
Invite to interview
Download CVCV
Overview
Technical skills
Timeline
Roles
Overview
AI data-quality engineer at a junior level specializing in small-scale CSV-based evaluation and dataset validation workflows. The strongest proven skill is building simple, reliable data validation and evaluation scripts as shown by validate_data.py and evaluate_responses.py. There is no evidence of model training, experiment tracking, MLOps, or scalable production infrastructure.
Phone
Technical skills
Python
AI/ML
Datasets
LLM Evaluation
AI/ML
LLM Apps
MLOps
LLMOps
Computer Vision
DevOps
GitHub
Netlify
Databases
MS SQL
Analytics
Microsoft Excel
Marketing
Amplitude
Timeline
Data Service Specialist
•
Middle
RWS
•
Full-Time
Performed structured evaluation and quality assessment of AI voice conversation outputs against project guidelines. Compared and ranked conversational responses for accuracy, consistency, and overall quality. Focused on dataset and output validation to support client delivery.
Reviewer III (Certified Atomic Action Labeller)
•
Middle
Atlas Capture LLC
•
Full-Time
Validated and audited large-scale AI datasets for structural accuracy, quality benchmarks, and policy compliance. Carried out atomic action labeling workflows to support training for generative AI and computer vision models. Ensured labeling outputs met defined quality standards during review cycles.
Computer Vision
Project Development Officer II (Technical Support)
•
Middle
Department of Social Welfare & Development (DSWD)
•
Full-Time
Provided technical support, data verification, and administrative coordination for regional office projects. Managed project documentation and stakeholder communications. Updated and maintained regional databases to support ongoing operational work.
Digital Content & Chat Moderator
•
Middle
Stream Smoothie
•
Full-Time
Monitored live digital communications to ensure compliance with platform standards and safety rules. Maintained appropriate client tone during chat sessions. Escalated or adjusted moderation actions to align with operational guidelines.
Data Researcher
•
Middle
Tektos Ecosystem, Ltd.
•
Full-Time
Conducted targeted market research and compiled structured datasets for operational and business intelligence reporting. Organized findings into usable formats for further analysis. Supported decision-making by preparing dataset-ready research outputs.
Junior AI/ML Engineer
Confidence: Medium Data-centric
AI data-quality engineer at a junior level specializing in small-scale CSV-based evaluation and dataset validation workflows. The strongest proven skill is building simple, reliable data validation and evaluation scripts as shown by validate_data.py and evaluate_responses.py. There is no evidence of model training, experiment tracking, MLOps, or scalable production infrastructure.
Model Architecture & Training
How well models are designed and trained
Not evidenced in public code
Data Pipeline & Feature Engineering
2/10
How data is prepared for models
Basic data-validation and CSV processing logic is implemented, including duplicate detection and required-field checks; simple parsing and row-level checks only.
Evidence
ai-evaluation-data-quality-toolkit/src/validate_data.py: validate_data - CSV parsing, duplicate ID detection, required-field presence checks and per-row validation messages
ai-evaluation-data-quality-toolkit/src/evaluate_responses.py: evaluate_responses - reads CSV rows and extracts fields for downstream aggregation
Experimentation & Evaluation
2/10
How results are measured and tested
Minimal evaluation/reporting functionality via console prints and simple aggregate counts; no experiment tracking, metrics dashboarding, or reproducible experiment harness.
Evidence
ai-evaluation-data-quality-toolkit/src/evaluate_responses.py: evaluate_responses - prints evaluation summary, preference counts and per-row reasons
MLOps & Deployment
How models are shipped to production
Not evidenced in public code
Computational Efficiency
How efficiently computing resources are used
Not evidenced in public code
Research Depth & Innovation
Depth of research and new ideas
Not evidenced in public code
Expertise
Conversational AI & Chatbots• Junior
Technologies
AI/ML
LLM Apps
Netlify
Computer Vision• since 2025
Datasets
GitHub
Python• mentioned only
Recommendations
- Add unit tests and simple CI to validate behavior and prevent regressions for the validation and evaluation scripts.
- Integrate experiment tracking or structured reporting (for example W&B, MLflow, or structured JSON/YAML outputs) to capture evaluation runs and metrics reproducibly.
- Expand the data pipeline with schema-driven validation (for example Great Expectations or custom schema checks), support for multiple input formats, and automated report generation.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories:
Intern DevOps Engineer
Confidence: Low Generalist
AI data specialist (junior level) focusing on LLM evaluation and dataset quality assurance. The strongest proven skill is LLM evaluation and localization testing, evidenced by the Tagalog-English LLM Evaluation Framework README and its atomic rubric descriptions. There is no human-authored operational or infrastructure code publicly present, so CI/CD, IaC, Kubernetes, observability and incident artifacts are not evidenced.
CI/CD Pipelines
Automated build and deploy
Not evidenced in public code
Infrastructure as Code
Managing servers with code
Not evidenced in public code
Containerization & Orchestration
Working with containers
Not evidenced in public code
Observability & Monitoring
Watching system health
Not evidenced in public code
Reliability & Incident Response
Keeping systems up
Not evidenced in public code
Cloud & Cost Optimization
Smart use of the cloud
Not evidenced in public code
Industries
Artificial Intelligence• Intern
Data & Analytics• Intern
Technologies
MLOps
LLMOps
Recommendations
- Lead development of LLM evaluation and benchmarking suites and atomic rubrics for localization and instruction adherence.
- Design dataset annotation and validation pipelines and automation for QA workflows in collaboration with MLOps engineers.
- Document and operationalize evaluation metrics and repeatable benchmark runs so platform engineers can integrate them into CI pipelines.
- Contribute to build-out of RLHF data curation processes and tooling for high-quality training/eval datasets.
Repositories
The developer's experience in this domain has been verified based on AI analysis of the following repositories: