Abhijit K.

Your finance team is losing 10–20 hours a week to work that should run itself. I rebuild those workflows into systems that just run — invoice processing, reconciliations, reports, and the data pipelines feeding them — all handling themselves without anyone babysitting them. 8+ years inside financial services, building data pipelines and automation for investment and fund data at enterprise scale. 5.0 on Upwork. I've seen exactly what breaks when this is done badly — the fragile macro nobody can touch, the pipeline that silently drops rows, the dashboard that goes stale — so I build it clean the first time. What I fix: Manual invoice processing → PDF extraction, validation, straight into your accounting system Reconciliations that eat days — bank vs. ledger, AR/AP, automatic exception flagging Reports rebuilt every week from files scattered across tools Dashboards that die because refreshing them is somebody's job Messy, inconsistent data with no single source of truth Endless copy-paste between QuickBooks, SharePoint, Excel, and Drive What I build: Data engineering & pipelines End-to-end data pipelines: ingest → clean → transform → deliver, fully automated Medallion architecture (bronze/silver/gold) on Databricks and Azure — raw data in, trusted reporting data out Data validation, quality checks, and error handling built in, so bad data gets caught, not shipped Scalable pipelines that hold up as volume grows, not just on today's sample file Automation Full Microsoft 365 automation — Power Automate, Office Scripts, Power Query, SharePoint working as one system, not scattered parts Power Automate flows: invoice data from email → accounting entry, fully hands-off Python pipelines that clean, reconcile, and deliver reports with zero human touch n8n webhook workflows wiring 400+ apps together with real branching logic AI classification that sorts thousands of records in seconds Extraction & reporting Robust PDF/table extraction (pdfplumber, pytesseract, AI-based) — the messy documents most tools choke on Power BI / Looker Studio dashboards on live data — no manual refresh, ever VBA and Apps Script for spreadsheet-level automation Tech stack: Python (Pandas, pdfplumber, pytesseract), Azure, Databricks, Power Automate, Power Query, Office Scripts, n8n, VBA, Apps Script, Power BI, Looker Studio, SharePoint, QuickBooks API, REST APIs. How I work: 30-min discovery call → architecture proposal → build & test → walkthrough + documentation → post-delivery support. Clean, documented code designed to still be running a year from now. Best fit for: teams with a real broken workflow — or messy data — ready for a permanent fix. I'll take small Excel/Sheets jobs too if they're the first piece of something bigger — happy to do a small build first so you can see how I work before you commit. Tell me what your team does by hand every week. I'll reply with exactly how I'd automate it, what it saves, and what it costs Abhijit