Looking for a job

Ashwin Badamikar

@ashwinbadamikar

Message

I build production AI agents and data platforms that improve research, analytics, and decision-making.

United States

Message

What I'm looking for

I'm looking to build production AI agents, RAG systems, and data platforms where evaluation, reliability, and measurable business outcomes matter.

At Aosenuma, I built StakeAI, a production multi-agent research automation system that gave analysts natural-language access to more than 500 daily ESG document outputs. I also developed RapidAI evaluation workflows that reduced hallucinations by 22% and model-drift detection time by 95%.

Previously at TurnOn, I built Kafka streaming pipelines for an agricultural IoT analytics platform handling more than 100,000 daily sensor events, improving throughput by 20%. I also delivered governed Airflow, Snowflake, and dbt pipelines and Power BI analytics that improved forecasting accuracy and accelerated client decisions.

I'm continuing to build agentic AI and data products, including HIPAA-compliant healthcare systems, SEC financial-data automation, and multimodal RAG applications. My work combines Python, LangGraph, retrieval systems, evaluation workflows, and production data infrastructure.

Experience

Work history, roles, and key accomplishments

AI Engineer Intern

Aosenuma Inc.

Jan 2026 - Jun 2026 (5 months)

Built StakeAI, a production multi-agent research automation system in LangGraph that orchestrated entity extraction, knowledge graph generation, and structured report writing, giving analysts natural language access to over 500 daily document outputs. Reduced hallucination rates by 22% and cut model drift detection time by 95% on RapidAI, an agentic evaluation pipeline built across 4 LLM configura

LangGraph Multi agent orchestration RAG Prompt Engineering Evals Pinecone Helicone LLM Agentic AI Python

Software Engineer

TurnOn

Jan 2023 - Jul 2024 (1 year 6 months)

Delivered Bhanush, an agricultural IoT analytics platform, by building Kafka streaming pipelines partitioned by device ID and time window, improving throughput by 20% across more than 100,000 daily sensor events. Accelerated client decisions by 15% on Smart Build Automation by shipping ETL pipelines with Pydantic schema validation and real-time Power BI dashboards. Built a single auditable source

Kafka Streaming ETL Pydantic Power BI Airflow Snowflake DBT SQL Python

Research Assistant

JSS Academy of Technical Education

Sep 2022 - May 2023 (8 months)

Earned national recognition from the Karnataka State Council for Science and Technology for leading Cross Lingual Transfer Learning for NLI across 15 language pairs on the XNLI benchmark, choosing multilingual BERT fine tuning. Improved reproducibility across a 4-person research team by reviewing code and evaluation pipelines weekly and translating findings into peer-reviewed reports.

BERT (Language Model)Transfer Learning NLI XNLI Research Code Review Evaluation Pipelines