Making complex manufactured products manageable.

Hi, I'm Kalyani. I work in discrete manufacturing, on products that come in thousands of variants: the configuration rules, bills of materials and engineering changes that decide whether a configured order can be quoted, built and shipped on time.

My work spans access control devices at Reliance Jio, HVAC and power equipment at Lumina Datamatics, and configure-to-order and make-to-order modular office space product lines across 7 brands at Herman Miller, the company behind the Eames Lounge Chair and the Aeron, which has ~1.2 million product configurations.

At Herman Miller I also worked on the decisions that shape a product line: which SKUs stay orderable from launch to end of life, how dealer pricing changes, and how to get new products to market faster. We cut time to market by 22%. What interests me most is the whole portfolio: which variants earn their place, which quietly cost more than they sell, and how a product line stays easy to buy as it grows.

7+years in configure-to-order manufacturing
120+engineering changes released a month
30%fewer quoting errors

What I bring

5,000+configurable products modelled

Product configuration and CPQ

Constraint-based models, 150% BOMs and routings in the Infor LN Product Configurator and Infor CPQ. Quoting errors cut by 30% for North American B2B dealers and OEMs.

120+engineering changes a month

Engineering change and PLM

ECR to ECO to ECN at 99.5% on time, with cycle time cut by 45%. Effectivity, impact analysis and open orders handled, not hoped for.

7brands, NPI to end of life

Product and portfolio decisions

SKU orderability, lifecycle and dealer pricing across 7 brands, 22% faster time-to-market, and an AI product taken from customer interviews to paying pilots.

Work

Articles, labs and a glossary

Most of what decides whether a configured order ships on time is invisible: a missing constraint, a stale dealer price book, a variant nobody remembered to retire. This is where I write about that layer, and test what I think I know.

Labs

Two working models I built: one for the rules behind a single product, one for the decisions behind a whole portfolio.

Explore both labs
The Rules Lab showing a configured lock, its bill of materials and routing
Configuration lab

Configure-to-Order Rules Lab

A commercial mortise lock modelled the way configure-to-order manufacturers really do it: the rule set, a 150% BOM with conditional lines, routing and price build-up, plus a regression suite.

  • 8 constraints that guide the user instead of failing at order entry
  • A revision cut-in and a finish phase-out, driven by order date
  • 18 test configurations that run before every rule change
The Portfolio Lab showing keep, retire and reprice decisions with their impact on net contribution
Portfolio lab

Variant Portfolio Lab

A task chair with 1,166,400 possible combinations, the same order of complexity as the Aeron. Decide which options to keep, retire or reprice, and see what each choice does to the business.

  • The full variant space and its long tail, counted live
  • Keep, retire and reprice decisions with their reasoning
  • Net contribution, cost to keep and valid configurations, before and after
How I think

A few things I believe about configurable products

My depth is in configuration rules, BOM structure and change control. I'm not an expert in every corner of PLM, and these are positions, not laws. Tell me where you'd push back.

If it can be quoted, it must be buildable

A configurator is a promise. Every constraint exists so the quote, the MBOM and the routing agree before the order reaches the plant.

Every option has a running cost

Adding a variant takes a meeting. Keeping it costs price-book columns, rules, tests and stock for years. Portfolios should be pruned as deliberately as they grow.

An ECO isn't done when it's approved

It's done when effectivity, stock disposition and every open order on the where-used list have been accounted for.

Fix the rule, not the order

Fallout is a defect report on the rule set. Correct the order, then fix the constraint and save the failure as a test configuration.

Career path

Work experience

  1. Product Owner / Product Manager

    Feb 2026 to May 2026

    Bantrly, Plano, TX

    • Took a new AI B2B product from 15+ customer and stakeholder interviews to an MVP in 3 months, with zero release rollbacks and 2 paying pilot customers.
    • Raised pilot user activation from 45% to 80% within 6 weeks by defining usage KPIs and fixing the top 3 drop-off points.
  2. Product Information Analyst, North America

    Oct 2020 to Dec 2023

    MillerKnoll, formerly Herman Miller, Bangalore, India

    • Owned the configuration rules behind Infor CPQ and order management for configure-to-order and make-to-order product lines across 7 brands, releasing 120+ engineering changes a month at 99.5% on time and cutting cycle time by 45%.
    • Reduced quoting errors by 30% as single point owner of order fallout for North American B2B dealers and OEMs, leading root-cause analysis with engineering, manufacturing and sales teams.
    • Built constraint-based models in the Infor LN Product Configurator (PCF), translating engineering drawings and EBOMs into generic (150%) manufacturing BOMs and routings with effectivity.
    • Governed SKU orderability and lifecycle states from NPI to obsolescence, and improved new portfolio time to market by 22% with SQL and Power BI launch readiness dashboards.
  3. Product Data Analyst, APAC and EMEA

    Feb 2020 to Oct 2020

    Herman Miller, Bangalore, India

    • Improved operational efficiency by 34% as owner of color, material and finish (CMF) master data for the APAC and EMEA product portfolio, harmonizing records across Infor LN and Infor SyteLine during the ERP data migration.

    Herman Miller makes the Eames Lounge Chair and the Aeron, a chair in MoMA's permanent collection with ~1.2 million product configurations.

  4. Founder

    Sep 2018 to Jan 2020

    Gogetters, Navi Mumbai, India

    • Launched a self-funded D2C brand, growing monthly revenue 3x at a 38% gross margin over 15 months by negotiating better terms with 12 suppliers and cutting acquisition cost by 30%.
  5. Senior Analyst

    Jun 2016 to Aug 2018

    Lumina Datamatics, Navi Mumbai, India

    • Built rule-based variant models for 5,000+ configurable HVAC, motor and power-equipment products, and cut production configuration errors by 30% by validating every rule change before release.
  6. Configuration Associate (Contract)

    Feb 2015 to May 2016

    Reliance Jio, Navi Mumbai, India

    • Maintained option rules for 300+ access controller devices (HID, Sterna) and processed 50+ engineering changes a month in PLM at 99% on time with zero escapes.
Education

Education

Northeastern University
Master of Science, Jan 2024 to Dec 2025

Information Systems

Northeastern University, San Jose, CA. GPA 3.87 of 4.0. Teaching assistant for Product and Project Management, and for User Experience and Design Testing.

MIT Pune
Bachelor of Engineering, 2010 to 2014

Electronics and Telecommunications

MIT Pune (University of Pune), India.

Let's talk shop

Have a complex product challenge, a configurator that needs taming, or a product line with more variants than anyone can hold in their head? I'd love to hear about it.

I'm open to product configuration, engineering change and product line roles in discrete manufacturing. Based in the Dallas-Fort Worth Metroplex and ready to relocate anywhere in the United States. Authorized to work in the US.