Resume · Updated July 2026

Praneeth Paikray

Solutions Architect - AI · GenAI Platform Specialist · AI/ML Systems

Bengaluru, India · LinkedIn · GitHub · Medium · Hugging Face

Summary

AI Solutions Architect and GenAI platform specialist with nine years of experience designing enterprise AI products across staffing, financial services, enterprise IT, and data platforms. Builds production LLM applications, agent systems, RAG services, evaluation workflows, ML platforms, and recommendation systems. Current work focuses on Databricks AI, MLflow evaluation and tracing, Agent Bricks, Genie, Databricks Apps, Vector Search, governance, and reusable AI engineering workflows.

Experience

Databricks

Solutions Architect - AI · Bengaluru, India

  • Build production-oriented architecture patterns across MLflow evaluation and tracing, Agent Bricks, Genie, Databricks Apps, Vector Search, Unity Catalog, and serverless workloads.
  • Contribute to the open-source ai-dev-kit, including reusable Databricks skills and notebook workflows for AI-assisted development with Genie Code.
  • Design and deliver enterprise AI workshops, solution architectures, and hands-on demonstrations for classical ML, GenAI, and agent use cases.
  • Create reusable assets for agent evaluation, self-improving AI workflows, real-time agents, retrieval systems, and Databricks application development.

ManpowerGroup

Senior Generative AI Specialist · Bengaluru, India

  • Architected an agentic invoice processing system on Azure using LlamaIndex workflows, reaching 93% extraction accuracy, reducing processing time by 75%, and handling more than 50,000 documents per month.
  • Led delivery of an enterprise GenAI platform with evaluation gates, safety guardrails, prompt registry, and RAG services for more than 500 recruiter and sales users.
  • Built a compensation analytics platform on AWS SageMaker with vLLM-optimized Phi-3 to extract, cluster, and review salary signals by skill and job family.
  • Engineered an AI resume screening system on AWS EMR using fine-tuned embeddings, reducing time-to-fill by 21% and producing $700K in cost savings through Workday integration.

Fidelity Investments

Senior Data Scientist · Bengaluru, India

  • Developed an employee churn model with a 77% F1 score by combining external economic indicators, internal performance metrics, and employee network signals.
  • Designed AI and ML workflows for model development, evaluation, and deployment across Azure ML, MLflow, and internal data platforms.
  • Worked with business, engineering, and risk stakeholders to turn model outputs into reviewable decision workflows.

Fidelity Investments

Data Scientist · Bengaluru, India

  • Developed a conversational analytics platform using Azure Cognitive Services and a custom BERT-based model for real-time sentiment analysis, reducing manager review time by 30%.
  • Built reusable NLP components for text classification, evaluation, reporting, customer feedback analysis, and operational decision support.

Dell

Data Science Developer · Bengaluru, India

  • Developed a hybrid semantic and keyword knowledge recommendation system that reduced support ticket inflow by 11%.
  • Built NLP and search workflows for knowledge discovery, issue triage, self-service enablement, and production evaluation.

Tata Consultancy Services

Systems Engineer · Bengaluru, India

  • Built sentiment analysis workflows for CRM teams and Python and SQL pipelines for data extraction, reporting automation, and campaign analytics.

Selected public work

CLEAR-S Eval Harness ↗

A layered evaluation system for agent outputs, trajectories, safety, latency, and cost across prompts, frameworks, and models.

Databricks AI Intern ↗

An autonomous AI and ML engineering loop that researches, trains, measures, reproduces, deploys, and benchmarks what it builds.

LayoutScribe ↗

A Python package for converting PDF, PowerPoint, and Word documents into Markdown and layout-aware JSON using multimodal models.

Real-Time Agent APIs ↗

A FastAPI and WebSocket reference implementation for streaming agent events, tool activity, approvals, failures, and completion.

Skills

Languages
Python, SQL, PySpark, Bash
Databricks
MLflow, Agent Bricks, Genie, Databricks Apps, Vector Search, Unity Catalog, Lakehouse Monitoring, Lakeflow, Model Serving
GenAI and agents
RAG, agent orchestration, LlamaIndex, LangChain, LangGraph, CrewAI, DSPy, prompt engineering, evaluations, traces, guardrails
ML and NLP
LLMs, SLMs, transformers, BERT, embeddings, re-ranking, recommendation systems, forecasting, optimization, sentiment analysis
Cloud and tools
AWS, Bedrock, SageMaker, EMR, Azure ML, Azure Functions, FastAPI, PyTorch, scikit-learn, vLLM, Snowflake

Education

MTech in Data Science

Birla Institute of Technology and Science, Pilani · 2021 - 2023

BTech in Electrical Engineering

Odisha University of Technology and Research · 2013 - 2017

Writing, speaking, and service

  • Author of technical essays on Databricks application development, agent evaluation, model architecture, and agent frameworks through Towards AI and Medium.
  • Speaker and workshop lead for enterprise AI sessions covering MLflow, Agent Bricks, Genie, real-time agents, APIs, MCP, and agent evaluation.
  • Publication: Invoice Information Extraction: Methods and Performance Evaluation, October 2025.
  • Databricks certification credential ↗
  • ML Engineer with Omdena, analyzing satellite imagery for wheat crop classification and region detection, June to August 2019.