Sangram
Keshari
Ghose
Building production data and AI systems · Data Engineer
I build the systems behind decisions: ETL pipelines that hold up against messy production data, forecasting models validated on RMSE, MAE, and MAPE, and Agentic AI workflows that cut token cost without giving up answer quality. Four industry internships, 50,000+ records modelled, and reporting cycles cut by 60%.
Open to internships and full-time roles · Based in India · Remote or relocation.
Highlights and achievements
Shipped work and verified credentials, not adjectives.
Industry internships
Shipped work at Mindpex (Agentic AI solutions), Frost & Sullivan, Infinite Computer Solutions, and Zoho, spanning analytics, pipelines, and customer-facing AI delivery.
Agentic AI and solutions
Designed multi-step LLM workflows, optimized prompts and context, and translated customer requirements into AI solutions for workforce-retention platforms.
Microsoft Fabric
Fabric Data Engineer Associate, building and managing end-to-end data solutions on Microsoft’s unified analytics platform.
Google Cloud
Associate Cloud Engineer, designing, deploying, and operating secure, scalable solutions on Google Cloud Platform.
Deploytual
Deploy intelligence. Any data. Anywhere. An AI analytics platform with AutoML, natural-language querying, and one-click executive reports.
WorkforceIQ
End-to-end workforce analytics pipeline with attrition prediction, clustering, and sentiment analysis across more than 50,000 records.
About me
What I build, where I have built it, and what I am aiming at next.
I am Sangram Keshari Ghose, a data engineer who builds pipelines, forecasting systems, and Agentic AI workflows against real customer problems rather than tidy sample data.
My experience spans Mindpex (Agentic AI and solutions), Frost & Sullivan (market analytics and forecasting), Infinite Computer Solutions (enterprise analytics platforms), and Zoho (data preparation products). Each role paired hands-on engineering with the delivery and stakeholder work that gets a system used.
My work runs from ingestion and modelling through forecasting and LLM orchestration, all the way to the last mile: making the result something a team actually depends on.
- Data Engineer
- Forward Deployed Engineer
- Analytics Engineer
How I work
Four rules I hold to whether the system is a script or a platform.
Idempotent pipelines
Re-runs should be safe. Stages write clear outputs so failures are recoverable without side effects.
Data contracts
Validate schemas and quality at boundaries. Bad data should fail fast, not silently pollute downstream models.
Smallest usable slice
Ship the thinnest path that proves value, then harden. Prefer working increments over perfect plans.
Document for the next engineer
README, runbooks, and clear naming so someone else can operate and extend the system without a meeting.
Skills and tooling
The stack that takes data from raw ingest to a decision someone can act on.
Programming and data
Python and SQL are where most of the work happens, from ingestion scripts to analytical queries.
AI, ML, and agentic systems
Forecasting models and retrieval-grounded agents built to survive real inputs and real budgets.
Data engineering
Modelling, quality gates, and transformations that stay reliable across repeated runs.
Cloud, APIs, and integration
Services, containers, and integrations that make a pipeline usable by other systems.
Visualization and BI
Reporting that lands the answer in seconds rather than after a spreadsheet detour.
Solutions and delivery
Scoping, documentation, and stakeholder work that turn a requirement into a shipped system.
Experience
Four internships across Agentic AI, market analytics, enterprise data, and product.
AI and Solutions Engineering Intern (Agentic AI)
Mindpex
- Designed and developed Agentic AI workflows using LLMs, contextual inputs, and multi-step orchestration to support Mindpex’s predictive workforce-retention platform, built around more than 47 behavioral signals.
- Optimized ICP modeling, prompts, and LLM context, substantially reducing token consumption while preserving the relevance and quality of AI-generated outputs.
- Translated customer business and technical requirements into AI-driven solution approaches, supporting HR data integration, workforce-risk analysis, and 6-12 month predictive attrition intelligence.
- Contributed to solution scoping, technical documentation, service integration planning, and AI solution implementation, collaborating with customers and product and technical teams to resolve bottlenecks and refine customer-specific solutions.
Data and Analytics Intern
Frost & Sullivan
- Analyzed large-scale industry datasets using Python, Pandas, and SQL, supporting market intelligence, trend analysis, and demand forecasting across two industry sectors.
- Assisted in developing SARIMAX and Prophet forecasting models while performing data cleaning, feature engineering, and exploratory data analysis (EDA) to improve the quality and reliability of forecasting datasets.
- Evaluated forecasting models using RMSE, MAE, and MAPE, contributing to the validation of more than 10 forecasting scenarios and preparing analytical outputs for internal research reports and consulting deliverables.
- Collaborated with a six-member analytics team to interpret market trends, prepare business reports, and translate analytical findings into data-driven insights that supported client presentations and strategic research projects.
Software Engineer Intern
Infinite Computer Solutions
- Developed analytics-focused microservices and automated KPI dashboards using Node.js, Couchbase (N1QL), and REST APIs, delivering real-time workforce analytics for more than 20 stakeholders and reducing report turnaround time by 60%.
- Engineered ETL-style data pipelines from Couchbase, automating recurring reporting and eliminating approximately four hours of manual effort per week.
- Contributed to migration of enterprise data from IBM DB2 to Couchbase using a pub-sub architecture.
Summer Intern
Zoho · DataPrep 2.0
- Explored data preparation concepts including cleaning, transformation, and validation on Zoho DataPrep 2.0.
- Gained hands-on experience with ETL workflows, data profiling, and schema mapping through guided tasks and product exploration.
- Assisted in feature testing, dataset validation, and identification of data quality issues to support product development.
Projects
Two flagship builds, each with a full case study and open source code.
Deploytual
An AI-powered analytics platform that unifies data connectivity, natural-language querying, automated machine learning, and one-click reporting into a single, deployable engine.
Natural-language querying
Plain-English questions converted into SQL or Pandas automatically.
AutoML engine
Anomaly detection, Prophet forecasting, and clustering without ML expertise.
AI data cleaning
Missing-value alerts, outlier detection, and one-click fixes.
Executive reports
Boardroom-ready PDFs with AI-written summaries and charts.
WorkforceIQ
End-to-end analytics pipeline ingesting more than 50,000 employee records and 10,000 call transcripts, with attrition prediction, clustering, and sentiment analysis.
Attrition prediction
Random Forest on 50K+ records at 91%+ accuracy.
Employee segmentation
K-Means clustering with four behavioral personas including a high-risk segment.
Sentiment analysis
VADER scoring of 10K transcripts with ~92% label agreement.
Automated reporting
Power Automate cut manual HR reporting turnaround by 60%.
Education
B.Tech, Computer Science and Engineering
Gandhi Institute of Engineering and Technology University
Gunupur, Odisha, India
Certifications
Seven verifiable credentials across cloud, data, and AI. Every badge links to its issuer.
Associate Cloud Engineer
Deploying, managing, and monitoring cloud solutions on GCP.
Agentic AI Builder
AI-powered applications, agentic workflows, and RAG pipelines.
Fabric Data Engineer Associate
Data analytics solutions and pipelines with Microsoft Fabric.
SAP Business Data Cloud
Analytics, data management, and enterprise data solutions.
Analytics Cloud 2025 Professional
Data modeling, visualization, advanced analytics, and ML.
Machine Learning Professional
Classification, regression, clustering, and feature importance.
Data Engineering Professional
Data access, transformations, multi-dataset workflows, and data processing.
Let’s build something
Roles, collaborations, or a quick technical chat. I reply within a day.
Internship or full-time role? Here is the fastest way in.
Send the role, team, and start window. I reply within a day with availability, relevant work, and a short technical walkthrough if useful.
Three to six month data and analytics internships, full-time hours, remote or on-site in India.
Permanent positions on data platform, analytics, or customer-facing engineering teams.
Available immediately, with no notice period to serve for either an internship or a full-time offer.