About

An engineer who starts with your problem, not a tool.

I'm Shreyas Pachpute, an AI Product Engineer based in Surat, India. I am the company: one name, one person who owns the whole system, from the first conversation about what is broken to the fixes after launch.

Over two years at Commercient Sync I turned manual support, sales and data work into AI agents that run inside a live .NET product, replaced a $12k-a-month third-party bot, and served open-weight models on the company's own GPUs. At ZYLITIX I own the AI backend of DTAXO, a tax-prep platform for CPA firms: document intelligence, workpaper generation and AI review.

I am technology-agnostic by design. The stack is chosen by the problem and learned as fast as the problem requires. What stays constant is the method: understand the real workflow, decide what needs to exist, build the smallest thing that proves it, integrate it where people already work, and measure what changed.

Right now I am building two projects in the open: Dispatch, an AI operations team for businesses that run on email, PDFs and spreadsheets, and Roundtable, a multi-agent proposal team that argues its way to a defensible proposal. I also write about this work on Medium.

Credentials

Proof, on record.

  1. AWS Certified Machine Learning – Specialty (MLS-C01)

    Amazon Web Services

    Validates the ability to design, build, train, tune, deploy and maintain ML solutions on AWS, and to choose and justify the right approach for a business problem.

    Verify
  2. Amazon ML Summer School 2023

    Amazon (India) · 2023

    Selective, test-gated program taught by Amazon scientists: four weekends of applied ML modules with live Q&A from senior applied scientists.

    Verify

Toolbox

I don't collect technologies. I collect ways to solve problems.

Click a tool with a number to see what I built with it.

AI
models, serving, retrieval, tuning
AGENTS
tool use, planning loops, guardrails
AUTOMATION
schedulers, triggers, human-in-the-loop
WORKFLOWS
state machines, queues, retries
APIS
design, integration, contracts
BACKEND
services, auth, background jobs
FRONTEND
interfaces people actually use
DATA
extraction, structuring, storage
CLOUD
deploy, scale, observe
DEV TOOLS
whatever shortens the loop

Lab

Small systems, running.

Three working miniatures of the patterns behind the real products: an agent loop, a document extractor and a job queue with retries. They run in your browser.

Open the lab

Contact

Let's talk.