Skip to content
Aswin Ram K--:--:--IST
Open to Data / AI roles — remote

Aswin Ram K — Data Analyst& AI Product Builder.

I turn messy data into decisions — and build the AI products that ship them.

See my work Résumé
Records analysed
0K+Records analysed
Dashboards shipped
0+Dashboards shipped
ML model accuracy
0%ML model accuracy
Tiruchirappalli, Tamil Nadu, IndiaScroll

01About

Open to Data Science / AI Engineer roles

Analyst foundation, AI builder ambition.

I'm a data analyst and AI product builder in Tiruchirappalli. Since 2024 I've turned 60,000+ records — rental markets, credit, customers, operations — into dashboards and models people use to decide.

Now I build the AI layer on top of that work: a retrieval system that answers from your own documents, and a bridge that lets an assistant query Power BI directly. I know what clean, trustworthy data looks like, so the AI I build answers from it instead of guessing.

Aswin Ram K
Aswin Ram KTiruchirappalli
Based in
Tiruchirappalli, Tamil Nadu, India · IST
Education
B.E. in Computer Science and EngineeringMookambigai College of Engineering · 2021 — 2025
Building now
DocMind, a RAG system, and a Power BI MCP bridge
Languages
English (professional), Tamil (native)
Certified
Python for Data Analytics, Data Science with R, Machine Learning Specialization (in progress)

02Selected work

01 / 04

Systems I'm building, and work already in use.

AI systems in progress, and the dashboards and automation clients run on today. Every project opens into a full case study.

BuildingAI · Retrieval-augmented generation

DocMind — a RAG system that answers from your documents

Ask questions of your own PDFs and reports, and get answers that cite the exact passages they came from.

Read the case study Follow the build
docmind · chat with your docs Building
What were the key credit risks flagged in the Q3 report?
retrieved credit_report_q3.pdf gst_invoices.csv risk_policy.docx
Three risks stood out: rising credit exposure in the South region[1], a portion of invoices breaching credit limits[2], and delayed GST reconciliation on high-value orders[3].
Ask anything about your documents…

Illustrative preview — sample conversation

02Selected work

02 / 04
BuildingAI · Business intelligence

Power BI MCP

An MCP bridge that lets an AI assistant query a Power BI model directly and answer business questions with the real figures.

One source of truth: the assistant answers from the governed model, not from a copy of the data.

Read the case studySolo build — MCP server, tool design and query layer · 2026
assistant · connected to Power BI over MCP Building
Which region drove net sales last quarter?
mcp · list_measures mcp · run_dax
EVALUATE SUMMARIZECOLUMNS(
  Region[Name], "Net sales", [Net Sales])

South led, with 38% of net sales — from the model itself.

South38%
West27%
North21%
East14%

Illustrative preview — sample data

02Selected work

03 / 04
ShippedBusiness intelligence · Client work

Papers · enterprise dashboard

One interactive dashboard in place of a folder of spreadsheets, so leadership sees the health of the business in one place.

One trusted view of the business instead of a folder of spreadsheets.

Read the case studyFor Papers
papers · enterprise dashboard Live
Business overview

Revenue

₹3.6 Cr

+6%

Orders

1,980

+3%

Margin

21.9%

−0.4

Revenue by month

Margin trend

One definition per KPI — every tile, chart and filter reads the same model.

Illustrative preview — sample data

02Selected work

04 / 04
ShippedBusiness intelligence · AI · Automation

Production company systems

Sales, credit and purchase data brought into one model with self-refreshing Power BI dashboards, a RAG assistant that answers from the company's own documents, and a data-entry flow that checks every record before it is written.

Reporting runs off one model instead of a pile of exports — 40% faster.

Read the case studyFor a garment production company · 2025
production company · sales & credit Auto-refresh
›››

Sales by region

South
West
North
East

South · top customers

Retail partner A₹48.2 L
Retail partner B₹31.7 L
Distributor C₹22.9 L

Invoice · Retail partner A

InvoiceINV-10482
Amount₹2,48,600
Credit used82% of limit
StatusDue in 6 days

Refreshed automatically — no manual exports

Illustrative preview — sample data

02Selected work

Earlier work

Where it started.

Analytics projects from 2024 and 2025 — credit analytics and rental pricing. Measured outcomes, each with a full write-up.

  1. Credit & sales intelligenceBusiness intelligence · Client work40%faster reporting cycle8K+invoices analysed35%faster dashboards
  2. European rental pricingData analysis · Market research50K+listings analysed18%regional price variance found60%less time on data prep

03Experience

05 roles · 2024 — now

The work, on the record.

Freelance consulting, project work and internships — real datasets, real stakeholders, measured outcomes.

Self-employed

Jul 2025 — PresentFreelanceTamil Nadu, India · Remote

Data & Business Intelligence Consultant

  • Built 5+ enterprise Power BI dashboards for credit, sales and purchase analytics, cutting the reporting cycle by 40%.
  • Analysed 8,000+ invoices, GST records, credit limits and purchase orders to surface credit-exposure risk by region and business unit.
  • Designed and delivered a UI-driven enterprise dashboard for Papers that replaced a folder of spreadsheets.
  • Built dashboards, a RAG assistant over the company's own documents, and automated, validated data entry for a garment production company.
  • Built recruitment cards and hiring dashboards for private companies in Chennai.
  • Handling payment data for power-sector businesses: reconciling transactions and reporting payment status.
faster reporting
0%faster reporting
invoices analysed
0K+invoices analysed
faster dashboards
0%faster dashboards
Read the case study

Power BI · DAX · SQL · Power Query · Python · RAG

Production company — garment manufacturing

2025Freelance projectTamil Nadu, India · Remote

BI & AI Consultant (freelance project)

  • Built Power BI sales, credit and purchase dashboards on one data model, with automated refresh and drill-down from a region to a single invoice.
  • Built a RAG assistant that answers questions from the company's own documents and shows the source of every answer.
  • Automated data entry: a structured template with validation of required fields, formats and duplicates before anything is saved.
faster reporting
0%faster reporting
Read the case study

Power BI · DAX · Power Query · Python · RAG

VDart Inc.

Mar 2025 — May 2025InternshipTamil Nadu, India

Business Intelligence Analyst

  • Developed 5+ interactive Power BI dashboards and SQL reports on customer behaviour across 5,000+ records.
  • Ran full ETL and data validation with SQL and Power Query, holding 99% data accuracy for business-critical reporting.
  • Exploratory analysis surfaced customer segments and trends that helped reduce churn by 12%.
  • Automated 10+ recurring reporting workflows, saving about 25 hours of manual work each month.
data accuracy
0%data accuracy
churn reduced
0%churn reduced
saved / month
0hsaved / month

Power BI · SQL · Power Query · EDA · Excel

European rental market project

Jul 2024 — Jan 2025FreelanceRemote

Market Research & Pricing Analyst

  • Ran end-to-end analysis on 50,000+ European rental listings with Python (pandas, NumPy) and SQL.
  • Exploratory analysis revealed an 18% pricing variance across regions and pinpointed high-demand neighbourhoods.
  • Automated the data pipelines, cutting data-prep time by 60% and enabling weekly trend monitoring.
listings analysed
0K+listings analysed
faster data prep
0%faster data prep
price variance found
0%price variance found
Read the case study

Python · pandas · NumPy · SQL · Data visualisation

NIT Trichy

Mar 2024 — May 2024ML internshipTiruchirappalli, Tamil Nadu

Applied ML Engineer

  • Built a face-recognition attendance system with Python, OpenCV and supervised learning, reaching 92% recognition accuracy.
  • Implemented image preprocessing and real-time detection that cut false positives by 20%.
  • Trained and compared SVM and KNN models with cross-validation on precision, recall and F1.
  • Automated attendance for 100+ users, removing manual logging errors.
model accuracy
0%model accuracy
test images
0+test images
admin time saved
0%admin time saved

Python · OpenCV · scikit-learn · SVM / KNN

04Capabilities

The toolkit behind the work.

A map, not a list. Pick a domain to see the tools I use there and how deep I go.

40+skills across 6 domains

05Clients

06 engagements

Proof of work, bound in a book.

What I delivered for each client, and what changed. Pick a name or turn the pages.

Vol. 03CE

Chennai enterprise client

Enterprise · Credit-exposure analytics

8K+invoices analysed Open the book
Vol. 02PA

Papers

Enterprise dashboard

Delivered Open the book
Vol. 01PC

Production company

Garment production · Dashboards, RAG & data entry

40%faster reporting Open the book

Swipe or use the arrows · tap a card to open

06Contact

Let's build something worth shipping.

Open to Data Science and AI Engineer roles — remote, or relocating — and to freelance work. If you have a problem worth solving with data or AI, write to me.

aswinram312004@gmail.com

Available for new work · it's --:--:--IST in Tiruchirappalli