Pavithran M
Responsible AI Engineer
Chennai,TN
Summary
Seasoned Responsible AI engineer and data scientist with nearly 6 years across Accenture, Hexaware, Cognizant, and TCS, specializing in AI governance, NLP, and generative AI for enterprise systems; lead end-to-end assessments and audits against 140+ metrics spanning fairness, transparency, explainability, robustness, and sustainability with actionable risk mitigations; design and operationalize governance controls aligned to regulatory expectations and internal standards, integrating decision logs, red-team reviews, and guardrails; build and productionize LLM/RAG systems with LangChain and LlamaIndex, operationalizing explainability and relevance scoring via DeepEval-style evaluations and custom scorecards; architect agentic A2A orchestration with safety boundaries, rollback strategies, and full-stack observability via OpenTelemetry; deploy on AWS (SageMaker, Bedrock, Lambda) with secure data pipelines on PostgreSQL and Snowflake and CI/CD best practices; strong in Python and SQL, Transformers, SpaCy, and Hugging Face, with deep experience in bias/toxicity evaluation and robustness testing against hallucination and jailbreak attacks; SME and mentor driving cross-functional alignment, measurable governance outcomes, and continuous auditability across the ML lifecycle.
Overview
6
6
years of professional experience
3
3
Languages
Work History
Responsible AI Engineer
Accenture
01.2025 - Current
- Engineered end-to-end Responsible AI frameworks with policy-as-code guardrails, automated lineage, and metric gates mapped to EU AI Act risk tiers and internal standards.
- Ran continuous audits across 140+ metrics (fairness, transparency, explainability, robustness, sustainability) with bias tests like disparate impact and specificity parity, producing mitigation plans and waivers.
- Built production LLM/RAG services with LangChain and LlamaIndex, adding evaluator agents for relevance, faithfulness, and toxicity; integrated IBM RAI/LIME for model transparency.
- Designed agentic A2A orchestration with safety boundaries, role-based capabilities, rollback strategies, and circuit-breakers; instrumented full telemetry using OpenTelemetry.
- Implemented secure MLOps on AWS (SageMaker, Bedrock, Lambda) with CI/CD, dataset versioning, and reproducible pipelines over PostgreSQL and Snowflake.
- Led red-team exercises (prompt injection, jailbreaks, data leakage) and codified decision logs and approval workflows to enforce accountable model releases.
- Built metric scorecards and dashboards for exec and audit consumption; drove continuous improvement of accuracy, fairness, and robustness SLAs.
- Mentored engineers and delivered org-wide training and workshops on Responsible AI and agentic patterns, establishing an internal center of excellence.
- Led the design and implementation of responsible AI frameworks, ensuring compliance with ethical standards and industry regulations.
- Collaborated with cross-functional teams to develop and refine AI models, enhancing system transparency and user trust.
- Conducted thorough risk assessments for AI initiatives, identifying potential biases and recommending corrective actions to mitigate impact.
- Developed training programs on responsible AI practices, empowering team members to integrate ethical considerations into their projects.
Generative AI Developer
Hexaware Technologies
02.2024 - 01.2025
- As a Generative AI Developer at Hexaware Technologies in India, I harnessed cutting-edge AI and machine learning technologies to develop innovative solutions for our clients. My role involved creating advanced generative models and leveraging state-of-the-art tools and platforms to autonomously generate content such as images, text, and music.
- Developed generative models using GANs, OpenAI's GPT models, and diffusion models for diverse content generation tasks.
- Integrated LangChain and LlamaIndex to build and deploy sophisticated AI applications with seamless data retrieval and interaction capabilities.
- Leveraged AWS Bedrock models to enhance the scalability and efficiency of AI-driven solutions.
- Collaborated with cross-functional teams to design and implement custom generative AI solutions tailored to client needs.
- Conducted extensive research and experiments to optimize model performance, focusing on reducing latency and improving content quality.
- Contributed to the development of new AI products and services, including fine-tuning LLMs for domain-specific applications and creating end-to-end AI pipelines.
Associate Data Scientist
Cognizant Technology Solutions
11.2021 - 09.2022
- Analyzed large datasets using SQL and Python to deliver actionable insights for data-driven decision-making.
- Developed and maintained interactive dashboards and reports in POWER BI and Tableau.
- Collaborated with data engineering teams to optimize pipelines using Snowflake and Apache Spark.
- Implemented statistical models and data visualizations to enhance business intelligence efforts.
- Conducted rigorous testing and validation of models, ensuring accuracy and reliability.
- Automated data engineering workflows using Snowflake, AWS, and Python scripts.
NLP Engineer
Tata Consultancy Services
10.2018 - 10.2021
- Designed and implemented NLP algorithms for text processing, sentiment analysis, and language modelling.
- Developed and fine-tuned models using libraries such as NLTK, SpaCy, and Transformer-based models like BERT.
- Worked on multiple client projects, delivering customized NLP solutions to enhance user experiences and automate text analysis.
- Integrated NLP models into production environments, ensuring scalability and performance.
- Collaborated with data scientists and software engineers to deploy NLP solutions in cloud environments, including AWS.
- Production Support:
- Managed deploying applications using JAMF, BBCA, and Active Directory tools, successfully deploying apps and security patches to over 30,000 machines.
- Generated routine reports to ensure the integrity of the production environment.
- Resolved critical tickets promptly, minimising disruptions and addressing infrastructure issues.
Education
Master of Science - Data Science And Business Analytics
University of Plymouth
Plymouth, United Kingdom 04.2001 -
Bachelor Of Engineering - Information Technology
Anna University
Chennai, India 04.2001 -
Skills
Proficient in Python programming
Machine Learning: OPENAI, AWS Bedrock model, Scikit-learn, Transformer models (BERT, GPT)
NLP Tools: NLTK, OpenAI GPT, Hugging Face Transformers
Data Visualization: POWER BI, Tableau, Matplotlib, Seaborn
Agentic AI (A2A protocol)
Langraph,CrewAI, DeepEval
Responsible AI frameworks
Ethical AI compliance
AI model development
Bias mitigation
Risk assessment
AI performance analysis
AI governance
Team collaboration
Decision-making
Problem-solving
Data governance
Agile methodologies
Interests
Photography, Reading, Swimming
Awards
Employee of the month from Client, Received SPOT recognition Awards for faster phase in work
Courses
- Data Analysis 2023 Microsoft and Linkedin
- Research journal published on Climatology in Mines using FM and RSSI 2018 IJRTCC
Links
pavithran-mohan-b76bb6152
Personal info
- Date of birth: 21 September 1996
- Place of birth: Chennai
- Nationality: Indian
- Driving licence: yes
Timeline
Responsible AI Engineer
Accenture
01.2025 - Current
Generative AI Developer
Hexaware Technologies
02.2024 - 01.2025
Associate Data Scientist
Cognizant Technology Solutions
11.2021 - 09.2022
NLP Engineer
Tata Consultancy Services
10.2018 - 10.2021
Master of Science - Data Science And Business Analytics
University of Plymouth
04.2001 -
Bachelor Of Engineering - Information Technology
Anna University
04.2001 -