Software engineer · alsoFintech backends
Software engineer who builds financial systems that reconcile to the .
Core craft is backend and distributed systems — reconciliation engines, event-driven services, and loan onboarding platforms in Python, Django, PostgreSQL, and AWS. I build the frontend interfaces those systems need too, so the dashboards are as reliable as the data behind them. Based in Jamshedpur, Jharkhand, India; remote-ready.
3.5+
Years professional experience
30+
Projects shipped
800+
DSA problems solved
4
Fintech & startups
Experience
~/ship_log
4 orgs · 1 live
// career.deployments
Production timeline
From first commits to live reconciliation — every role as a shipped mission.
$ git log --oneline --deploys
CacheTech Advisor Solutions
APR 2025 — PRESENT
./cachetech
Solvative Pvt Ltd
OCT 2024 — JAN 2025
./solvative
Markytics Consulting Private Limited
JAN 2023 — JUL 2024
./markytics
Incrivelsoft Pvt Ltd
DEC 2021 — JUN 2022
./incrivelsoft
bootstrap · not counted in companies
DC-00Decartem (Bahadur and Bisht Technologies Pvt Ltd)
▸ 01Worked on Attendo, an ERP and communication solution, alongside a team of 20 interns.
~/recon_engine
Custodian-scale holdings, tax lots, and transactions that have to agree — with conflict detection and resolution workflows operators can trust.
// ledger.diff
sync · ok
matched ~87k lots · queued resolve
reconciliation pipeline
$ watch -n1 recon status
~15.7k
Holdings / Schwab run
~93k
Tax lots validated
46.15s
~87k tax-lot writes
0
Failed SNS publishes
Problem
Advisor households hold positions across multiple custodians — Schwab, Fidelity, private funds — and those books have to reconcile: holdings, tax lots, and transactions at scale. When they don't, operators need clear conflict detection and resolution workflows, not silent mismatch.
Approach
Reengineered core matching into a dedicated raw-SQL microservice, added event-driven SNS fan-out for household-level updates, migrated the data layer to psycopg3 (recon_v2), and built the UI for bulk relabeling, conflict resolution, Tax vs Daily Quote views, transaction mapping, and hybrid auto/manual workflows.
// recon_engine.evidence
What shipped and what held up
6 proof points · validated
Handled custodian-scale reconciliation runs — e.g. Schwab ~15.7k holdings, ~93k tax lots, ~10k transactions in a validated full run.
Tax-lot upsert completed in 46.15s for ~87k row writes during a Schwab validation run.
Automated SNS household fan-out with 0 failed publishes across validated Schwab and private-fund runs.
Shipped a psycopg3 migration (recon_v2) through test, staging, and production, and fixed a production multi-item resolve regression (PR #2020).
Built the UI/API for bulk relabeling, conflict resolution, hybrid auto/manual workflows, Tax vs Daily Quote views, and a transaction mapping tab.
Reengineered and deployed the reconciliation platform in production across four distinct wealth-management and custodial firms, each with its own reconciliation configuration and conflict-resolution rules.
~/loan_onboarding
End-to-end lending onboarding system — built, demoed to banks, won a client, then customized for production use.
// verify.pipeline
5 / 5 · clear
application → sanction → disbursal
go-to-market arc
$ trace journey --from build
5+
Banks / lenders demoed
5
Verification sources
1
Client won & customized
4
Stages to production
Problem
Rural and semi-urban bank customers often had to visit branches physically for loan applications, sanctions, and disbursals — waiting in queues for hours, only to have paperwork errors send them back to repeat the process. There was no digital path for a customer to apply, get verified, and track a loan from application through disbursal without an in-person branch visit.
Approach
Built a digital, end-to-end loan onboarding system (Django, PostgreSQL, React Native, React.js, AWS) that lets applicants apply, get verified, and track their loan status remotely — with real-time verification against CIBIL, PAN, Aadhaar, bank statements, and ITR replacing manual document checks, cutting the paperwork back-and-forth that caused repeat branch visits.
// loan_onboarding.journey
How the system reached production
Build → Demo → Win → Customize
Built the loan onboarding system end to end (Django, PostgreSQL, React Native, React.js, AWS), integrating real-time verification against CIBIL, PAN, Aadhaar, bank statements, and ITR.
Supported product demos to 5+ banksand lending companies alongside the company's leadership and sales team — sometimes personally walking prospects through the product via screen share and live narration, other times providing technical screen-share support while leadership led the pitch.
Contributed to winning Meet Finance Pvt Ltd as a client.
Customized the system's workflow / business logic and UI / brandingto match Meet Finance's specific requirements.
Projects
A focused sample from 30+ shipped projects — more case files below the featured studies above.
~/selected_work
3 case files · indexed
// work.archive
Case file index
Professional builds and freelance ships — each with its own signal.
$ ls case_files/ --sort=signal
Four internal systems built solo for Monginis, working directly with the Head of IT — customer data management, inventory management, production planning, and cost management. Owned everything from architecture through deployment on each.
// solo.ownership_pipeline
5 stages · end-to-end
PW-07
./compliance
Backend systems for a client's cloud compliance-checking platform — scraped compliance data for providers like AWS and Azure, and validated infrastructure configurations against regulatory requirements.
private · no public artifact
PW-05
./decent-services
Freelance service-marketplace build — customer-facing site plus admin tooling for managing listings and requests.
Skills
~/competitive_coding
5 hosts · online
// problems.solved
800+
Across contest platforms — open a host to inspect the profile.
$cat platforms.json | jq '.hosts'
~/engineering_stack
6 modules · loaded
// toolchain.manifest
Languages through systems thinking — the craft behind the case studies above.
$ tree skills/ --depth 1
01
./lang
Languages
3 entries
02
./backend
Backend
5 entries
03
./frontend
Frontend
4 entries
04
./data
Databases & Caching
3 entries
05
./infra
Cloud & DevOps
4 entries
06
./systems
Systems & Concepts
5 entries
Contact
Software engineering first — reconciliation systems, distributed backends, and the interfaces they need. Remote or hybrid.
Start a conversation