AI Automation · Toronto

Your business runs on manual work. It doesn't have to.

I'm a senior software engineer who builds practical automation for small and mid-sized businesses — invoice processing, reconciliation, reporting, and the operational glue in between. Real systems, wired into your real tools. Not demos.

Book a free intro call See what I build Taking 1–2 clients this quarter

Services

What I build

01

Finance & accounting automation

Invoice intake, data extraction, reconciliation, month-end reporting — the repetitive work that eats your team's week. Built with a practicing CPA in the loop, so the output matches how books are actually kept. e.g. PDF invoices → extracted line items → matched against your ledger → exceptions flagged for review

02

AI assistants wired into operations

Document Q&A, drafting, intake triage, internal knowledge search — assistants that answer from your documents and hand off to a human when they're not sure. e.g. client emails auto-drafted from engagement history, queued for approval

03

Systems integration & data plumbing

The unglamorous part that makes everything else work: connecting your tools, cleaning the data flow between them, and building the APIs and event pipelines that keep it reliable. e.g. CRM → accounting → reporting, one source of truth, no re-keying

04

Automation audit

A fixed-scope engagement: I map where your team's hours actually go, rank the automation opportunities by ROI, and hand you a concrete build plan — whether or not I'm the one who builds it. 2 weeks · written report + prioritized roadmap

About

Why me

I've spent ten years building the systems enterprises depend on — event-driven backends, healthcare data platforms, payment-scale integrations — at PointClickCare, The Home Depot, GE Aviation, and Novantas. Java, Spring Boot, Kafka, AWS and Azure: the boring, reliable stack that keeps running at 3 a.m.

Finance is home turf. I work alongside my wife, a CPA and Director of Finance, on accounting automation — which means every workflow I build is reviewed by someone who closes books for a living, not someone who skimmed an accounting blog.

I'm pragmatic about AI. Some problems need an LLM. Most need clean data plumbing, a good rule engine, and one well-placed model call. I'll tell you which is which before you spend anything.

Process

How an engagement works

WEEK 0

Discovery call

Free, 30 minutes. You describe the manual work that's costing you; I tell you honestly whether it's automatable and roughly what it'd take.

WEEKS 1–2

Pilot

One narrow workflow, end to end, running against real data. You see actual output before committing to anything bigger. Fixed price, agreed up front.

ONGOING

Build & handoff

We expand what works. Everything is documented, monitored, and yours — no lock-in, no dependency on me to keep the lights on.

Fit

Is this you?

Good fit

  • An SMB or professional-services firm drowning in manual finance/ops work
  • You have a specific, painful workflow in mind — or suspect several
  • Your data lives in real systems (or spreadsheets) and you can grant access
  • You want something that works, maintained, over something flashy

Not a fit

  • You want an "AI strategy deck" — I build things, not slideware
  • You need a full-time hire (I take a small number of engagements alongside my day job)
  • The workflow changes weekly — automation needs a stable target

Contact

Let's talk about the work that's eating your week.

Email me with a sentence or two about the workflow on your mind. I'll reply within a day or two with an honest take — and if it's worth a call, we'll set one up.

akshay.bhandary9@gmail.com · Toronto, ON