Kartik Aneja
Kartik Aneja

AI/ML Platform Engineer in Boston. I build production AI systems, streaming data platforms, and the MLOps glue between them.

Now · updated September 2026

What I'm working on right now.

Inspired by nownownow.com. A snapshot, not a portfolio.

Focus

  • Looking for AI/ML Platform Engineer roles — open to the United States and any country that sponsors work visas.
  • Shipping the AI/ML platform-layer projects in /projects in public.

Building

  • evalstack: LLM eval framework with run + event diffs and a Next.js dashboard.
  • tracelens: drop-in LLM observability with SQLite or ClickHouse storage.
  • routerai: cost-aware LLM routing gateway — the govern layer completing the eval/observe/govern trilogy.
  • instaml: real-time feature store — YAML defs, DuckDB + Redis.
  • dataask: NL→SQL for founders + PMs, sqlglot-safety-gated against DuckDB.
  • lakehouseit: open-source modern data stack — Postgres → Debezium → Iceberg-style Parquet → dbt → DuckDB.
  • docubrain: PDF Q&A that cites the exact page an answer came from, enforced at the data-model level — or it doesn't answer at all.

Reading / learning

  • Eugene Yan's writing on LLM evaluation patterns.
  • Designing Data-Intensive Applications — re-read on streaming primitives.
  • Anthropic + OpenAI agent and tool-use cookbooks.

Open to

  • Full-time AI/ML platform, AI infrastructure, MLOps, or ML engineering roles.
  • Conversations with teams building production LLM systems or feature platforms.
  • United States or any country with work-visa sponsorship — remote-friendly. Book time or email.
This page changes month to month. If you're reading it long after the date above, the snapshot is stale — ping me for what's current.