Headline

Data Scientist @ EverestIMS | Building GenAI & LLM Systems | RAG, Fine-tuning, Agents | Production AI |

About

I build ML systems that run in production, not in notebooks. At EverestIMS, I designed and shipped a hybrid RAG pipeline (dense + sparse retrieval over FAISS and pgvector) that powers intelligent event management, root-cause analysis, and automated remedy recommendations for enterprise IT infrastructure. I also build event-driven data pipelines (Kafka, RabbitMQ, PostgreSQL) that cut alert noise and speed up anomaly detection at scale — and integrated an LLM into the product to auto-generate test scripts, cutting manual QA effort significantly. Before that, I spent 3+ years as a software engineer building the backend services and data infrastructure that my ML work now runs on. That's my edge: I don't hand off models — I own them from prototype to production. Outside work, I experiment in public: → buildresumenow.in — a free, ATS-friendly AI resume builder for the Indian job market → India's most complete 2025 startup funding dataset (887 deals, live on Kaggle with dashboards) → Deep dives into agent frameworks (Google ADK, Claude Agent SDK, Microsoft Agent Framework) Currently pursuing an M.S. in Data Science & ML (Woolf). Open to: Data Scientist / ML Engineer / AI Engineer roles. Teams & founders: I help with AI/LLM architecture decisions — what to build, what to skip, and how to ship fast. DM me if you're experimenting.

Location

Bengaluru

AI stack

claude code

deepseek

gemini

langgraph

openai

huggingface

Work experience

Data Scientist at EverestIMS Technologies Limited, 2026 - Present · 10 mos, - Designed and shipped a Hybrid RAG pipeline (dense + sparse retrieval over FAISS and pgvector) powering intelligent event management, root cause analysis, and automated remedy recommendations — improving incident triage speed and relevance of surfaced results. - Built event-driven data pipelines ingesting multi-source telemetry via Kafka and RabbitMQ into PostgreSQL, reducing alert noise and enabling faster, more accurate anomaly detection at scale. - Developed and optimized predictive analytics, clustering, and anomaly detection models ; integrated a large language model into the product framework for intelligent test-script generation, cutting manual QA effort significantly. - Drove end-to-end ML lifecycle ownership from use-case definition and prototyping to production deployment across multiple AI features in a fast-moving, cross-functional team. - Translated business requirements into demonstrable ML features through stakeholder prototypes and proof-of-concepts, accelerating adoption of data-driven workflows across the product.

Software Engineer at EverestIMS Technologies Limited, Oct 2022 - Feb 2026 · 3 yrs 5 mos, Bengaluru, - Built and maintained full-stack backend services (Python, PostgreSQL, MongoDB) and event-driven workflows using Kafka and RabbitMQ across distributed microservices for the NMS platform. - Developed and optimized REST APIs and data models, improving system reliability and data flow infrastructure that later became the backbone for ML and AIOps feature development. - Collaborated cross-functionally on early ML use-cases and data pipeline design, bridging software engineering and data science workflows.

Python Developer at Adaptive learning solutions, Jul 2021 - Oct 2022 · 1 yr 4 mos

Internship Trainee at Raam Group, Feb 2020 - Mar 2020 · 2 mos, Hyderabad, Telangana, India

Education

Masters, Data science & Machine learning, Woolf

Bachelor of Technology - BTech, Instrumentation Technology/Technician, VNR Vignana Jyothi Institute of Engineering and Technology (VNRVJIET)

Public links

github: https://github.com/pranay-dasari

pranay-dasari-portfolio.web.app: https://pranay-dasari-portfolio.web.app/