Jan 2025 – Present · 1 yr 5 mos · Pune, Maharashtra, India

• Led & Mentored: Headed a pod of three engineers, conducting 2‑person code reviews and weekly deep‑dive sessions to uphold code quality and accelerate skill growth.
• Architected Scalable ML Platform: Designed a Python‑based pipeline on AWS (EKS, Lambda, S3, SQS/SNS) processing 200M+ records, ensuring 99.9% availability and sub‑second inference.
• Modular & Extensible Codebase: Co‑developed builder–factory patterns and configuration‑driven modules for rapid feature onboarding—reducing new‑feature integration time by 40%.
• Automated A/B Testing Framework: Collaborated with an architect to implement AWS Step Functions, Lambda, and ECS tasks for end‑to‑end experiment orchestration—boosting developer productivity by 60%.
• Performance Optimization: Migrated ETL from Pandas to Polars for multi‑threaded transformations—shrinking batch windows from 24 to 4 hours—and provisioned concurrency to cut Lambda cold‑start latency by 60%.
• CI/CD & DevOps: Built automated pipelines with GitHub Actions and CloudFormation, enabling zero‑downtime UAT and production deployments.
• Cross‑Functional Collaboration: Partnered with data scientists, ML engineers, solution architects, and data engineers to translate complex requirements into robust, data‑driven solutions.
• Multi‑Modal AI Integration: Integrated RAG‑style inference, OpenAI Whisper & Amazon Transcribe for audio, and custom vision modules for image understanding in conversational workflows.
• Java Competency: Leveraged Java expertise to build complementary microservices, ensuring strict SLAs for long‑running, high‑throughput tasks.

May 2023 – Jan 2025 · 1 yr 8 mos · Pune, Maharashtra, India

• Architected & Delivered a personalized Generative AI chatbot from concept to production, leveraging LangChain for Custom Agent, Prompt Templates, and Output Parsers to handle diverse user intents.
• Engineered Dynamic Connectors enabling real‑time integrations with internal APIs and external data sources, ensuring seamless, context‑aware responses.
• Designed Multi‑Layer Safety Framework for GenAI applications—integrating SLMs, custom ML models, and guardrails—to block 99% of unsafe or off‑policy outputs, and implemented automated safety tests using Garak.
• Evaluated & Benchmarked a spectrum of LLMs, SLMs, and ML models to identify best‑fit solutions for enterprise use cases across NLP, summarization, and recommendation tasks.
• Developed Microservices Architecture using API Gateway, message passing (SQS/SNS), and circuit breaker patterns to achieve fault‑tolerant, loosely coupled services.
• Containerized & Orchestrated all components with Docker and Kubernetes on AWS EKS, maintaining 99.9% uptime and zero‑downtime deployments.
• Optimized Cost & Performance by tuning model invocation pipelines, reducing average response latency by 40% and cutting infrastructure costs by 25%.
• Collaborated Cross‑Functionally with data scientists, ML engineers, solution architects, and security teams to align on requirements, design reviews, and compliance standards.
• Mentored & Evangelized best practices in code quality, CI/CD (GitHub Actions, CloudFormation), and Agile workflows—accelerating feature delivery by 30% and enhancing team productivity.