AC / NYCContact ↗

Adhyyan “Adi” Chhabra

BUSINESS STRATEGY LEAD · BUILDER · NEW YORK, NY

The strategy person
who ships the tools.

I run commercial strategy for a $400M+ electrical-distribution business across two divisions. I also design and build the software that runs it.

LOAD A$400M+P&L owned
LOAD B97%quote-matching accuracy
LOAD C20,000+hours removed annually
LOAD D2%expansion in division profit
Adhyyan “Adi” ChhabraCED · Great Lakes & Elevator Divisions5+ years

INTERNAL SYSTEMS / BUILT AT CED

Software with a
balance-sheet address.

I built the systems below inside CED. They are company intellectual property, so I share the business context, what I built, and the results—not code, repositories, or live demos.

A

Five applications · 30+ profit centers

FlowSuite

DEPLOYED

ContextFive critical workflows—purchasing, obsolete inventory, vendor cost updates, sales risk, and category reporting—were slow, manual, and fragmented across profit centers.

What I builtI designed and built FlowSuite: five Python applications using pandas and openpyxl, with configurable business logic so each profit center could keep the operating rules it needed.

ResultPurchaseFlow cuts buyer cycles from hours to minutes across 20,000+ SKUs and 100+ manufacturers. WriteDownFlow removes roughly 8 hours of manual work per week for 50+ users. CostFlow processes a 100–150-file batch in under 5 minutes. SalesFlow cuts weekly reporting from an hour to under 5 minutes across 30+ profit centers.

B

Claude API · semantic search · Python

QuoteParser

DEPLOYED

ContextSales reps received handwritten and natural-language product requests that did not map cleanly to the internal SKU catalog, turning every quote into a manual search.

What I builtI built QuoteParser in Python with the Claude API and semantic search. It interprets free text, matches it to internal products, and assembles a finished quote for review.

ResultIt reaches 97% matching accuracy and saves roughly 10 minutes per quote for a team producing 30+ quotes a day.

C

Internal marketplace · Lovable · React

CED Yard

BUILT · LIVE PILOT

ContextSurplus and obsolete inventory sat idle or leaked to third-party marketplaces because CED locations could not see what other branches held.

What I builtI built CED Yard, an internal marketplace that lets locations discover and transact surplus inventory inside CED instead of sending it outside the company.

ResultI launched it in July 2026. In its first week, 175 active users moved $120K in product. It is now piloting across 100 locations ahead of a company-wide rollout to CED’s 800-location network.

D

Python · Flask · Playwright · Claude API

Competitive Price Intelligence

DEPLOYED

ContextCommercial teams were quoting without a current, like-for-like market reference from Home Depot, Lowe’s, and Menards.

What I builtI built a real-time competitive price intelligence system in Python, Flask, and Playwright. Claude reasoning matches non-identical retailer listings to CED SKUs.

ResultIt surfaces pricing deltas against quoted prices so the commercial team can review market position before acting.

E

Python · Claude API · Supabase · Streamlit

AI Purchasing Agent

BUILT · WORKING

ContextBuyers needed better decision support from demand and inventory data without giving up control of purchase orders.

What I builtI built an AI purchasing agent with Python, the Claude API, Supabase, and Streamlit. It reads demand and inventory signals and drafts purchase orders for buyer review.

ResultThe product is built and working. It produces review-ready purchase orders while keeping approval with the buyer.

INDEPENDENT BUILDS / OPEN SOURCE

Own IP.
Open circuit.

>_A

MediaMate

Claude API · Python · Claude Code

I built MediaMate to replace a $100/month SaaS subscription. It writes, schedules, and publishes social content autonomously at near-zero marginal cost.

REPO [FILL] ↗
>_B

Munch

Claude API · MCP

Recurring DoorDash orders were repetitive but still needed safety and budget controls. I built Munch to automate them end to end through MCP, with dry-run defaults, audit logs, dietary guardrails, and spend caps.

VIEW REPO ↗
>_C

Reservy

Claude API · Python · Claude Code

High-demand reservations require constant monitoring and fast action. I built Reservy to monitor availability and book through a reusable booking-adapter layer.

VIEW REPO ↗
>_D

Outbound GTM System

Instantly · Apollo · MillionVerifier

Cold outreach required too many disconnected steps. I built an end-to-end system for domain warm-up, list building, verification, and sequencing.

REPO [FILL] ↗

COMMERCIAL TRACK RECORD / FIELD RESULTS

The non-code half
is not a footnote.

~70%

less manual repricing

I built VBA and Excel automation for repricing across $15M+ in revenue, cutting the manual work by roughly 70%.

8–10%

profitability improvement

I built a SQL competitive-intelligence framework across 5,000+ SKUs, driving an 8–10% profitability improvement on targeted lines—about $750K annually.

$1.2M

secured each year

I built business cases and ROI models that secured $1.2M each year in vendor co-op and advertising funds.

15+

co-funded campaigns

I designed and ran 15+ co-funded vendor campaigns, producing 12% average revenue lift, 2% margin improvement, and $500K+ in annual profit.

$2M+

dead stock cleared

I built a demand-matching model that moved $2M+ in dead stock back into productive use.

$20M+

growth opportunities acted on

I built performance dashboards across 30+ locations, giving leadership one view of service levels, inventory turns, and margin leakage. The division acted on $20M+ in surfaced growth opportunities.

WALKTHROUGHS / RECORDING

Show the work.

REC ● STANDBY

Live demo video

Recording in progress

[FILL]
REC ● STANDBY

Architecture walkthrough video

Recording in progress

[FILL]

EDUCATION / FOUNDATION

Indiana University
Kelley School of Business

BS in Business — Business Analytics, Digital Tech Management, Economic Consulting.