Avik Mandal _{_{// Data Engineer → Data + AI Engineer}}

_{Bangalore, India • Shipping data platforms • Building with AI, building for AI}

`~/flow` _{how the work flows}

flowchart LR
    A[("SOURCES<br/>Kafka · APIs · S3")]
    B["INGEST<br/>Flink · Spark · Glue"]
    C["ORCHESTRATE<br/>Airflow 3.x · MWAA"]
    D[("LAKE<br/>Iceberg · S3 · MinIO")]
    E["QUERY<br/>Trino · Athena · DuckDB"]
    F(["AI LAYER<br/>LLM tooling · RAG · Bedrock"])
    G[["PRODUCTS<br/>dashboards · features · agents"]]

    A --> B --> C --> D --> E --> G
    D -.-> F -.-> G
    C -.-> F

    classDef ai fill:#2D1B4E,stroke:#8957E5,stroke-width:2px,color:#E6EDF3;
    classDef core fill:#161B22,stroke:#30363D,color:#E6EDF3;
    class F,G ai;
    class A,B,C,D,E core;

_{The solid path is where I’ve lived for a decade. The dotted path is where I’m building next.}

`~/focus` _{right now}

Platform work	Airflow 3.x · Kubernetes · Iceberg + Trino lakehouse patterns
AWS	MWAA · EMR · Glue · Athena · S3 · Lambda · ECS
AI in the loop	Using AI across the dev cycle — design, code, review, docs
AI in the product	Shipping features where AI measurably improves UX
Learning	AWS Data Engineer cert · Go for backend services · model serving

`~/projects` _pinned

local-infra-setup
_{Local dev infra with all services wired together.}
_{Shell · Docker · Makefile} airflow3-dags
_{Production-style DAG patterns exploring Airflow 3.x.}
_{Python · Airflow 3.x}

`~/stack` _{by layer}

_{ORCHESTRATION}
_AWS
_INFRA
_LANGUAGES
_DATA
_{AI / ML}	_{upskilling →}

`~/stats` _{github signal}

`~/talks` _{speaking & open source}

🎤 Speaking

_{Airflow DAG patterns at scale}
_{Internal tech talks — production-grade orchestration, failure modes, observability.}

_{Data platform case studies}
_{Lakehouse architecture, cost/perf trade-offs. Available on request.}

_{→ Invite me to speak}

🌱 Open Source

_{Airflow community}
_{Contributing around Airflow 3.x patterns, providers, and dev ergonomics.}

_{Local-dev tooling}
_{local-infra-setup — batteries-included dev stack.}

_{→ All repositories}

✍️ Writing

_{Production DAG patterns}
_{Notes on idempotency, backfills, and observability in Airflow.}

_{AI-augmented engineering}
_{What actually moves the needle in the daily dev loop.}

_{→ Posts on LinkedIn}

`~/ask-me-about`

Apache Airflow · Data pipeline design · Kubernetes for data · Data lake architecture · AWS data stack · AI-augmented development · Shipping AI features in data products

Avik Mandal // Data Engineer → Data + AI Engineer

~/flow how the work flows

~/focus right now

~/projects pinned

~/stack by layer

~/stats github signal

~/talks speaking & open source