Skip to content
View mitulpatel123's full-sized avatar
馃彔
Working from home
馃彔
Working from home

Block or report mitulpatel123

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don鈥檛 include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user鈥檚 behavior. Learn more about reporting abuse.

Report abuse
mitulpatel123/README.md
Mitul Patel - cloud, backend, and applied AI engineer

FlightOps AI Data Reliability LinkedIn Virginia

Engineering profile

I am an M.S. Information Technology student in Virginia, graduating in April 2027. I build cloud-ready backend, data, and applied AI systems where the engineering evidence is visible: typed contracts, failure handling, tests, observability, evaluation, and documented tradeoffs.

My 2027 focus is software engineering, backend/platform, cloud, data engineering, and applied AI/ML. Internship work authorization is coordinated through my university's CPT process and confirmed for each role before applying.

Shipped systems

FlightOps AI

FlightOps AI is an airline-operations intelligence API for retry-safe flight-event ingestion, explainable operational risk, and evaluated T-24h arrival-delay prediction.

official BTS data -> chronological evaluation -> portable model -> versioned API
Proof What is implemented
Live API Interactive OpenAPI documentation on Vercel
Reliability Idempotent ingestion, validation, health checks, and metrics
Engineering Eight automated tests, Docker, CI, architecture decisions, and scoped production deployment
Evaluated ML 90,000 official BTS records; future-period test; ROC-AUC 0.640; PR-AUC 0.335; recall 0.748; Brier 0.176
Model integrity Leakage-resistant T-24h contract, calibration, carrier/airport/time error slices, model card, and inspectable JSON inference

Stack: Python 路 FastAPI 路 Pydantic 路 SQL/SQLite 路 Docker 路 GitHub Actions 路 REST/OpenAPI 路 Vercel

Airline Data Reliability Lab

Airline Data Reliability Lab is an observable, quality-gated airline batch pipeline that records what loaded, what failed, and why.

CSV -> schema + business checks -> quality gate -> facts or quarantine -> API + metrics + report
Proof What is implemented
Data quality Required-schema checks, reason-coded row quarantine, and fail-closed batch thresholds
Reliability Deterministic fact keys make source retries idempotent; every run records its source SHA-256
PostgreSQL path SQLAlchemy schema, PostgreSQL 16 Compose environment, run/check/fact/quarantine tables
Verification Seven tests, passing CI, FastAPI evidence endpoints, Prometheus metrics, and a generated report

Stack: Python 路 FastAPI 路 PostgreSQL 路 SQLAlchemy 路 Docker Compose 路 GitHub Actions 路 Prometheus

Current focus

  1. Turn the two shipped systems into tailored resume bullets and role-specific application evidence.
  2. Confirm CPT dates and limits with the university DSO before accepting or starting any internship.
  3. Add AWS/Terraform depth only after the application pipeline is active and tracked.

Technical direction

Systems Data + AI Delivery
Python, FastAPI, REST, SQL Data contracts, baselines, evaluation Docker, CI/CD, cloud deployment
TypeScript, React Retrieval, embeddings, model cards Observability, failure handling, documentation

Working principles

  • Ship evidence, not keyword lists. Every featured skill should point to code, tests, a live system, or measured results.
  • Design for failure. Retries, duplicate events, invalid inputs, degraded dependencies, and rollback paths are part of the product.
  • Measure AI honestly. Start with a baseline; report calibration, latency, errors, and limitations before claiming improvement.
  • Keep provenance clear. External projects are learning references or attributed forks鈥攏ot cosmetically modified work presented as original.

Current signal

status: internship_application_ready
shipped_projects: 2
flagship: flightops-ai-v0.2
current_priority: tailored applications + targeted outreach
target_roles:
  - software engineering
  - backend / platform engineering
  - cloud / data engineering
  - applied AI / ML engineering

Open to 2026-2027 U.S. internship conversations in cloud, backend, data, and applied AI.

Pinned Loading

  1. flightops-ai flightops-ai Public

    Airline operations intelligence API with idempotent event ingestion, explainable delay-risk scoring, tests, Docker, and CI.

    Python

  2. airline-data-reliability-lab airline-data-reliability-lab Public

    Observable, quality-gated airline data pipeline with PostgreSQL, quarantine evidence, FastAPI, and metrics

    Python