MUKUL A. BHATTFDE
FORWARD DEPLOYED ENGINEERBENGALURU, IN

Mukul Anand Bhatt

I build backend.

I build production backend, automations, and integrations, then get on the call to make sure it actually ships.

Mukul Anand Bhatt
~ SHIPS ON THE CALL
30,000+
USERS REACHED
200+
COMPANIES REACHED
$1,000s
MRR SAVED FOR CLIENTS
01 /WHAT I BRING
01 / BUILD

Production from scratch.

Microservices, data pipelines, and third-party integrations that hold up under real load.

NodeTypeScriptPythonPostgresMongoRedis
02 / DEPLOY WITH CUSTOMERS

On the call, not behind it.

Client calls, live troubleshooting, custom solutions. Shipped while the customer watches it work.

client callslive debuggingcustom fixes
03 / OPTIMIZE

Cheaper, harder, faster.

Codebase optimizations that cut cost, raise reliability, and scale without drama.

cost ↓reliability ↑scale
02 /SELECTED WORK
04 case studies →

Built. Deployed. Owned.

01
FLAGSHIPSUPERJOIN · FDE

Data Connector: the core product

End-to-end owner of the engine syncing 34+ third-party sources into Sheets & Excel as one source of truth. Now automating it to run on autopilot. 30k+ users, 200+ companies.

02
AI-NATIVEAGENTIC DEV WORKFLOW

Claude Pickup

Non-technical testers file tickets → Claude drafts a PRD → I review → Claude builds → I test → merge. ~40% of issues resolved end-to-end. (This very site was built this way.)

03
GOV SCALEIRCTC · CONTRACT

Food-Safety Inspection Backend

Backend for India's national food-safety inspection platform. 1,000+ daily inspections with geo-location validation, at government scale.

04
ENTERPRISEKISNA · BYTIVE

Kisna Jewellers Integration

Microservice syncing MongoDB → MSSQL to feed Salesforce ingestion for an enterprise client, plus GoKwik + Reward Rally payment/rewards integration.

03 /HOW I WORK

The FDE loop.

One day on the job, on repeat. Scroll and walk it with me.

▼ SCROLL · ONE DAY IN THE LOOP
A DAY IN THE FDE LOOP · 5 STOPS, THEN REPEAT
01 SHOW UP02 DISCOVER03 BUILD04 OPTIMIZE05 AUTOMATE⟲ REPEAT
04 /MY CLAUDE SETUP· AI-NATIVE

I don't just use AI. I've wired it into production. Safely.

From Slack, I pull the full picture on any user or issue: logs, data, code, usage, billing, recordings. Every connection is read-only: it sees everything, changes nothing. Then Hermes turns the fix into a PR I just review.

ONE REAL SESSION · SCROLL TO REPLAY IT
claude · slack command center
 
▼ SCROLL TO RUN THE SESSION
THE ARCHITECTURE BEHIND IT
pulls full context · ALL READ-ONLYAny user · any issueSLACKcommand centerAWSlogsREAD-ONLYMongodataREAD-ONLYCodebaseREAD-ONLYFlexPriceusageREAD-ONLYStripebillingREAD-ONLYIntercomconvosREAD-ONLYPostHogrecordingsREAD-ONLYClaude · deep diagnosisroot cause across all of it · zero write accessHERMESACTION LAYERcode changes · pull logs · open PRsI review & merge
HOVER / TAP A CARD TO TRACE IT IN THE DIAGRAM
ONE COMMAND CENTER

I query any user or any issue straight from Slack.

FULL-STACK CONTEXT

Logs, data, code, usage, billing, conversations, recordings. All read-only by design.

SAFE TO POINT AT PROD

Deep root-cause analysis with zero write access. Nothing changes.

HERMES ACTS

Makes code changes, pulls logs, opens PRs. I review and merge.

AI-NATIVE, SECURITY-FIRST

The AI diagnoses and drafts. The human decides and ships.

05 /PROOF OF WORK· LIVE

Shipping, daily.

LIVE FROM GITHUB · mukul-anand-bhatt
897+CONTRIBUTIONS / YR
37DAY STREAK
MON
WED
FRI
AUG
SEP
OCT
NOV
DEC
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
LESSMOREREFRESHES EVERY 6H
06 /EXPERIENCE· 4 STOPS

Where I've shipped.

FEB 2026 - PRESENT01 / 04

Forward Deployed Engineer · Superjoin

End-to-end owner of the Data Connector, the engine syncing 34+ third-party sources into Sheets & Excel for 30k+ users across 200+ companies.

  • Leading the automation push so the connector runs on autopilot: self-healing instead of hand-held.
  • Join client calls to troubleshoot live and ship custom fixes on the spot.
  • Wired Claude into prod (read-only) for root-cause analysis straight from Slack.
Node.jsTypeScriptSheets APIAWSRediscase study →
JUN 2025 - JAN 202602 / 04

Backend Developer · bytive.in

Enterprise data integration for Kisna Jewellers: cross-database sync plus payments and rewards, without disrupting the existing stack.

  • Built the microservice syncing MongoDB → MSSQL on a reliable cadence, unblocking Salesforce ingestion.
  • Integrated GoKwik (payments) and Reward Rally (rewards) into the platform.
Node.jsMongoDBMSSQLSalesforcemicroservicescase study →
JUN 2025 - SEP 202503 / 04

Backend Developer · Contract · IRCTC

Backend for India's national food-safety inspection platform. Government scale, government scrutiny.

  • Designed the data flow to sustain 1,000+ inspections per day reliably.
  • Implemented geo-location validation so every inspection is verified at its claimed location.
Node.jsExpressPostgreSQLgeo-validationAWScase study →
JUN 2024 - JUL 202404 / 04

Data Engineer Intern · HERE Technologies

Data pipelines and processing at map-scale. My first taste of data that doesn't fit on one machine.

  • Built and maintained pipelines processing geospatial data at global map scale.
Pythondata pipelinesgeospatial
07 /STACK

The toolbox.

LANGUAGES
PythonTypeScriptJavaScriptSQL
BACKEND
Node.jsExpressPrisma
DATA / STORAGE
PostgreSQLMongoDBRedis
INTEGRATIONS
data pipelinesAPI connectorsSheets / Excel API
AI & AUTOMATION
OpenAI APIsLLM integrationagentic workflowsn8nprompt engineering
CLOUD / DEVOPS
AWSDockerCI/CD
ARCHITECTURE
REST APIsmicroservicesRBAC authsystem optimization