Python Developer and LLM Trainer

VINOTH KUMAR

Python Developer

+91-979**-***** *********.*@*****.*** CHENNAI, Tamil Nadu

LinkedIn: https://www.linkedin.com/in/thevinod1512/ Website: https://github.com/vinod523 PROFESSIONAL SUMMARY

Results-driven Python Developer with over 5 years of experience in building scalable applications, RESTful APIs, and full-stack web solutions. Recently expanded into the field of AI model training as a Coding LLM Trainer and Data Annotator, contributing to the fine-tuning of large language models for leading AI firms such as Outlier.ai, Alignerr, and SuperAnnotate. Skilled in Python, JavaScript, SQL, and frameworks like Django and Flask, with proven strengths in code evaluation, prompt engineering, and model alignment using RLHF methodologies. Adept at blending software engineering expertise with AI alignment tasks to support next-generation language technologies. EXPERIENCE

Coding LLM Trainer / AI Data Annotator, Alignerr, February 2025 - Present Reviewed and improved AI-generated code snippets, explanations, and debugging suggestions to align outputs with expert-level programming practices.

Annotated code tasks, logical reasoning chains, and multi-step problem-solving prompts for training and fine-tuning LLMs.

Participated in quality assurance of data labeling workflows, ensuring consistency and accuracy across diverse coding domains.

Provided structured feedback on prompt-response pairs to help improve alignment, coherence, and instructional effectiveness.

LLM Trainer / AI Data Annotator, SuperAnnotate, February 2025 - Present Contributed to fine-tuning and evaluation of large language models by reviewing AI-generated coding solutions and technical explanations.

Annotated and curated high-quality datasets focused on software development tasks, logical reasoning, and code comprehension.

Applied domain expertise in software engineering to identify edge cases, improve instruction-following, and align outputs with developer best practices.

Worked cross-functionally with QA and research teams to iterate on annotation protocols and feedback accuracy. Coding LLM Trainer, Outlier.ai, April 2024 - Present Delivered expert-level code reviews and evaluations for AI-generated programming solutions, improving model reasoning and output quality across Python, JavaScript, and other languages. Annotated complex programming tasks, debugging logic, and performance issues to support reinforcement learning and fine-tuning of large language models.

Simulated real-world developer workflows and problem-solving strategies to train LLMs for better alignment with professional coding standards.

Collaborated with AI researchers and engineering teams to iterate on prompt quality, code correctness benchmarks, and training data guidelines.

Python Developer, Fort Technologies, June 2019 - Present Mumbai

Software Development: Design, code, test, and debug software applications using Python. Python Programming: Proficient in writing clean, readable, and efficient Python code. Web Development: Familiarity with front-end technologies (HTML, CSS, JavaScript) and the ability to work on full-stack development projects.

Code Optimization: Debug and optimize code for performance and scalability, ensuring the efficiency and reliability of applications.

Database Integration: Integrate databases into applications and ensure proper data storage, retrieval, and manipulation.

API Development: Design and implement RESTful APIs for seamless integration with external services and applications.

Documentation: Create and maintain technical documentation for code, APIs, and software architecture. Testing: Conduct thorough testing of applications to identify and fix defects, ensuring the reliability and robustness of the software.

Frameworks: Experience with popular Python frameworks such as Django, Flask, or FastAPI. Junior Software Engineer, sanmina, June 2016 - April 2018 Chennai

Provided optimal project support to development teams in a 125-employee IT firm. Identified opportunities for product optimization.

Built and maintained web application for the accounting platform. Contributed boldly original or pragmatically simple ideas during weekly software design discussions. Problem-solved with various stakeholders.

EDUCATION

ELECTRICAL AND ELECTRONICS ENGINEERING

ANNA UNIVERSITY, Jun 2013

DIPLOMA IN ELECTRICAL AND ELECTRONICS ENGINEERING

BHAKTAVATCHALAM POLYTECHNIC COLLEGE, May 2010

SKILLS

Python, Java, JavaScript, C++, PHP HTML, CSS, SQL

Django, Flask, FastAPI LLM Code Review & Prompt Evaluation Reinforcement Learning with Human Feedback

(RLHF)

AI Model Alignment & Instruction Tuning

Annotation Tools: SuperAnnotate, Surge AI, Scale AI Test Case Design & Error Analysis HOBBIES AND INTERESTS

Tech Meetups and Conferences: Attend and participate in tech meetups and conferences to stay updated on industry trends and network with professionals.Coding Challenges: Engage in coding challenges on platforms to continually enhance problem-solving and coding skills.

VOLUNTEER AI PROJECTS

Marvel AI – CoTeacher (Volunteer Project)

Developing an AI assistant pre-configured to support educators with lesson planning, student engagement, and task automation.

Utilizes existing AI platform capabilities like actions, chat memory, and prompt suggestions to reduce setup complexity for users.

Marvel AI – Writing Feedback Generator (Volunteer Project) Building an AI tool that evaluates student writing based on rubric-aligned criteria and generates personalized feedback. Aims to reduce educator workload while improving feedback consistency and writing instruction outcomes. CERTIFICATIONS

Https://freecodecamp.org/certification

/fcc-a70cab64-8ada-4a69-88b5-b021ed9e10f5

/machine-learning-with-python-v7

Python Developer - https://micro1-

portaldata.s3.amazonaws.com/engineer-certificates

/1735532765-182ecd07-cdaf-4c6fa001-

d42738b2a0a7.jpg

RECENT AI ENGINEERING & BENCHMARKING WORK

Software Engineering Benchmark Contributor Project-based work

• Designed and validated software-engineering benchmark tasks across debugging, build systems, containerization, concurrency, security, and ML infrastructure.

• Created reference solutions, visible and hidden tests, executable grading checks, and reproducible Docker environments.

• Reviewed task solutions for correctness, test coverage, edge cases, environment reliability, and requirements adherence. Project Dynamo / Code Verifier V2

• Contributed to software-engineering evaluation tasks involving debugging, automated testing, build and dependency workflows, and containerized environments.

• Applied structured error analysis to identify logical defects, incomplete coverage, performance concerns, and reproducibility gaps.

Multi-Tenant Notification & Reminder Service

• Worked on backend workflows involving scheduled processing, transaction-related data, database consistency, auditability, and reliable multi-tenant operations.

• Applied idempotency controls, database-level unique constraints, retry handling, logging, and tests to reduce duplicate processing.

Otter Biomedical NER Benchmark

• Built and validated a biomedical named-entity-recognition benchmark with model training, development-set evaluation, and a measurable micro-F1 performance threshold.

• Created a reproducible submission package and verified the full Oracle and rubric workflow, including direct-execution and withheld-test checks.

• Resolved archive permission and packaging issues to ensure the deliverable ran reliably in the remote evaluation environment. TECHNICAL COMPETENCIES

• Backend & data: FastAPI, PostgreSQL, Redis, Celery, REST APIs, Pandas, NumPy, scheduled jobs, retry handling, idempotency, and database constraints.

• Machine learning & AI: PyTorch, TensorFlow/Keras, scikit-learn, XGBoost, LightGBM, Hugging Face Transformers, Spark MLlib, LLM evaluation, instruction tuning, and RLHF-related evaluation.

• AI quality & benchmarking: Code review, prompt evaluation, benchmark design, reference solutions, hidden and visible tests, executable grading checks, rubrics, test-case design, and error analysis.

• Platform & delivery: Docker, Kubernetes, AWS, Azure, GCP, GitHub Actions, CI/CD, automated testing, reproducible environments, and Git workflows.

• Collaboration: Branching, pull requests, code review, CI checks, issue linkage, documentation, and merge-conflict resolution. SELECTED PROJECT EXTENSION

AI-Powered Code Evaluator

• Built evaluation workflows for AI-generated code using task specifications, tests, rubrics, and repeatable execution environments.

• Focused on functional correctness, edge cases, test coverage, and the ability to reproduce results consistently. Roblox / Luau Evaluation Scenario

• Developed a Roblox/Luau evaluation harness with client-server behavior tests, a self-contained reference implementation, and rubric-aligned scoring criteria.

• Validated positive and negative outcomes in Roblox Studio using behavior-focused assertions, gameplay physics checks, and reproducible pass/fail evidence.

Fenrir C++ Fuzzing Benchmark

• Built a private C++ benchmark repository with fuzzing harnesses, seed corpora, ClusterFuzzLite configuration, and repeatable build, sanitizer, and regression-test validation.

Project Pluto / Harbor Task Authoring

• Authored containerized benchmark-task packages with task instructions, Docker environments, oracle scripts, programmatic tests, independent pre/post validation, and automated rubric, oracle, and difficulty checks.

Contact this candidate