now shippingmcoa-agent v0.1

Vigneshwar Lokoji

I build the rails that move data — and the agents that make sense of what arrives.

north starAspiring Chief of AI(CoAI)
AI NativeAgentic AI DeveloperPrompt EngineerAI Systems Builder
United States — open to remote
Get in touch GitHub LinkedIn
Vigneshwar Lokoji
where I've built
Texas A&MHigher EducationTCSIT ConsultingFirst AmericanFinTechTeamLeaseStaffingOwens CorningManufacturing
Live from the lab
commitspublic + private · last 12 weeks
… active days · … contributionslessmore
The work

Four hats, one mission.

I move between these depending on what the problem needs. They're more alike than the job titles suggest — every one of them is about turning information into something a team can act on.

Data Analyst

I translate messy data into decisions — dashboards, exploratory analysis, and the questions worth asking.

View projects

Data Engineer

I build the rails that move data — pipelines, warehouses, and the plumbing that makes everything else possible.

View projects

AI / ML Engineer

I take models from notebook to production — feature stores, training loops, and inference that scales.

View projects

Agentic AI

I design systems that reason and act — multi-agent orchestration, tool use, and evals you can trust.

View projects
AI Native

What I've been building with AI?

Tap the tile to break the glass — folders of everything I've shipped spill out and drift. Open one and the projects inside float up; click a project to bring it front and flip it for the README.

Artificial intelligence is the new electricity.
— Andrew Ng
Experience

Where I've worked.

A short list, on purpose. Tap any card to flip it for the full story.

Jan 2025 — May 2026 flip

Graduate Teaching & Research Assistant

Texas A&M University – Corpus Christi

Mathematics instruction, statistical-software migration, and educational analytics.

// full story↺ flip back

Graduate Teaching & Research Assistant

Texas A&M University – Corpus Christi · Jan 2025 — May 2026

  • Supplemental instructor for Pre-Calculus; ran dedicated tutoring + office hours; earned a 4.3 / 5 student satisfaction rating across the term.
  • Cross-disciplinary stats tutor: trained Marine Biology, Biological Sciences, and non-technical students on JMP and Jamovi for Statistics for Life; guided math majors through Calc I / II problem-solving in MATLAB.
  • Spearheaded the department's strategic migration from legacy JMP to Jamovi: authored 12 lecture modules spanning 8 classes plus a video-based training library to upskill the other TAs on the new stack.
  • Co-built a custom Canvas LMS REST API integration with fellow grad assistants to pull live student-performance + quiz data into an analytics DB; handled OAuth, pagination, and rate-limiting.
  • Modeled the daily quiz scores to identify concept-level knowledge gaps and surfaced them to faculty so lectures could be adjusted dynamically — drove a 37 % improvement in student retention and dropped the class dropout rate to 2 %.
  • Designed the underlying dashboards alongside faculty so end-of-class comprehension became a daily teaching signal, not a post-mortem.
PythonCanvas LMS APIJamoviJMPMATLABPower BI
Aug 2024 — Dec 2024 flip

IT Support Analyst

Texas A&M University – Corpus Christi

Authentication-log analysis, license auditing, and ServiceNow ticket automation.

// full story↺ flip back

IT Support Analyst

Texas A&M University – Corpus Christi · Aug 2024 — Dec 2024

  • Analyzed application-usage logs for 8 000+ users across Get Mobile, Islander Portal, and Duo Mobile to surface anomalous and high-cost access patterns.
  • Found that 23 % of Duo Mobile users were on the high-cost SMS / phone 2FA path; presented findings to management with a deprecation plan and authored the necessary transition knowledge bases — $29 000 in annual savings for the university.
  • Mined ServiceNow ticket queues to identify high-volume / low-complexity request categories; built and shipped automation for 17 % of total ticket volume, dropping manual handling on those categories to 0 %.
  • Audited Microsoft 365 and Intune license allocation across the user base; flagged unused / mis-provisioned licenses for the service-request team to reclaim.
  • Authored knowledge-base articles and self-service playbooks so students and staff could resolve common IT issues without filing a ticket.
ServiceNowPower BIPythonDuo MobileMicrosoft 365Intune
Jan 2022 — Jan 2024 flip

Data Engineer

Tata Consultancy Services (TCS)

Enterprise data integration, ITSM pipeline automation, and executive dashboarding.

// full story↺ flip back

Data Engineer

Tata Consultancy Services (TCS) · Jan 2022 — Jan 2024

  • Central data liaison across Service Request Management, ServiceNow developers, the SOC, Server, and Network teams — consolidated five disparate streams into a single unified executive dashboard via a custom EDW integration layer.
  • Wrote REST-API ingest pipelines that pulled ~79 000 historical ServiceNow records; parsed and flattened nested JSON / XML payloads into queryable relational tables.
  • Used data profiling on the ticket corpus to identify highly repetitive manual workflows; built event-driven automation that cut targeted-category manual handling from 38 % → 2 % (only edge-case tickets remained manual).
  • Stood up real-time monitoring dashboards in Power BI for server / system health, incoming alerts, service-request lifecycle, and SLA-breach detection — all fed by time-series telemetry pipelines.
  • Queried Microsoft 365 + Intune user-profile data to audit software license allocation across the org; helped the service-request team optimize distribution.
  • Operational transparency for management + proven ROI on the automation work directly contributed to the client renewing a 10-year customer-service contract with TCS.
PythonREST APIsJSON / XMLServiceNowPower BIAzure Entra
Feb 2021 — Dec 2021 flip

Data Analyst

First American Title Insurance Company

ETL pipelines, MySQL transformation, and Power BI reporting for property + claims data.

// full story↺ flip back

Data Analyst

First American Title Insurance Company · Feb 2021 — Dec 2021

  • Pulled raw property and claims records from distributed on-premise county servers using Python (pandas + sqlalchemy); cleansed and structured them in MySQL before handing the validated datasets off to QA.
  • Engineered the initial data landing zone — pushed raw extractions into a centralized S3 / Azure Blob bucket before transformation, moving the architecture from legacy ETL toward a modern ELT framework.
  • Wrote complex SQL with multi-table joins, subqueries, and window functions to flag discrepancies, isolate individual claim statuses, and aggregate group claims for the Process & Claims teams.
  • Built automated data-quality checks that gated downstream loads — duplicate-record and orphan-key detection, schema-evolution handling, date / null normalization — before anything hit the production warehouse.
  • Prepared and formatted the cleansed datasets for the downstream ETL team so Process and Claims operations always worked off validated data.
  • Built interactive Power BI dashboards that tracked new-record ingest volume, data-quality integrity, and customer-claim throughput end-to-end.
MySQLPythonPandasPower BIAWS S3
Education

Where I studied.

Texas A&M University – Corpus Christi

College of Science & Engineering

Master of Science in Data Science

Jan 2024 — May 2026
Corpus Christi, TX

Osmania University

Princeton College

Master of Science in Computer Science

Jan 2022 — Dec 2023
Hyderabad, India

Osmania University

Pragathi Mahavidyalaya

Bachelor of Science in Computer Science, Electronics & Mathematics

July 2017 — Dec 2020
Hyderabad, India
Research

Research.

Where the engineering meets the math — applied research work.

01
Applied research — undergraduate capstone·2023

Single-Image Super-Resolution of SAR Data Using Deep Learning

Deep CNN approach to upscaling synthetic-aperture radar imagery, recovering fine structure from low-resolution captures without paired ground truth at full scale.

Computer VisionSARDeep Learning
Stack

The tools I reach for.

Not exhaustive. The list I'd actually defend in an interview.

Languages
PythonTypeScriptJavaScriptRSQLBash
Data & Warehouse
PandasNumPyDuckDBBigQueryAirflowSupabase
AI / ML
PyTorchTensorFlowscikit-learnHugging FaceLangChainn8n
Cloud & Infra
GCP (Vertex AI)AWSDockerGitHub Actions
Databases
PostgreSQLBigQuerySupabaseRedis
Specializations
Computer VisionNLPBayesian StatisticsMultivariate AnalysisAgentic AI WorkflowsData Pipelines
Dispatches

Writing in public.

Notes from the lab, posted to LinkedIn — what shipped, what broke, and what I learned on the way.

LinkedIn2026

Shipping mcoa-agent v0.1 — a visual web-scraping agent

Gemini 1.5 Pro + Playwright, driven by screenshots instead of brittle selectors. Notes on what broke, what surprised me, and why GUI agents are harder than they look.

Agentic AIPlaywrightGemini
read
LinkedIn2026

From 38% manual to 2%: automating ITSM pipelines at TCS

A retrospective on event-driven ServiceNow automation — 79K tickets, webhook-fed EDW, and the dashboards that kept a 10-year contract alive.

Data EngineeringAutomationServiceNow
read
LinkedIn2026

What teaching statistics taught me about explaining models

Two semesters as a GTRA at TAMU-CC: if a pre-calc student can follow your reasoning, a stakeholder can follow your model. Communication is the real interface.

TeachingAnalyticsCommunication
read
Follow the feed
Stage time

Achievements & events.

Wins worth keeping a receipt for, plus the rooms I've sat in to learn from people doing the work.

Achievements

Achievement Coming Soon

TBD

Competitions, awards, and recognitions will appear here.

Events

Event Coming Soon

TBD

Conferences and events will be added once details are confirmed.

Contact

Let's build something.

I'm Vigneshwar — a data engineer and AI builder who has moved 79K tickets through automated pipelines, taught statistics to full lecture halls, and shipped agents that browse the web on their own. I work across the whole stack: the rails that move data and the models that make sense of it. If you're hiring, want to connect and discuss something, or just want to trade notes on evals — say hi or reach out below.

Send email LinkedIn