AI Product Leader & Senior AI/ML Engineer
Program Manager · Insight Global
yuktachikate14@gmail.comI work at the intersection of machine learning infrastructure and product strategy — building the evaluation layers, observability pipelines, and agentic architectures that let enterprise AI systems operate reliably in healthcare and finance, where failure has real consequences.
Over the past five years I've led cross-functional teams from 0-to-1 prototypes through production hardening — reducing hallucination risk, cutting inference latency, and shipping the safety rails that let organizations trust what their models say.
From Falling Water to Risk-Adaptive Verification
Yukta Chikate
A business and engineering playbook for building AI systems that verify, challenge, refine, abstain, or escalate before consequential actions are executed.
Chikate, Yukta
U.S. Patent Application No. 64/176,709 · Submitted October 14, 2026
Risk-Adaptive Verification Engine
Featured Research
Designed a runtime verification framework that treats the first AI output as a candidate action rather than a verdict, allocating additional verification according to risk before ACT, ABSTAIN, or ESCALATE.
Pilot-Centered Electronic Flight Bag
Research Prototype
Designed a map-first electronic flight bag prototype that keeps planning, weather, weight-and-balance, route state, and inspectable decision support close to a persistent spatial context.
July 8, 2026 — Present
Insight Global
Lead program reporting, analytics, and process improvement initiatives — building Power BI dashboards, defining KPIs, automating recurring reports, and translating operational data into clear insights for leadership decision-making.
Feb 2026 — July 2026
CVS Aetna — EXL
Led a 12-person cross-functional team across UX, data engineering, and frontend to ship enterprise AI and source-traceability interfaces, directing BigQuery semantic search and automated contract validation at scale.
June 2025 — Oct 2025
SnapLogic
Designed automated agentic test-case execution and deployed Prometheus/OpenTelemetry observability across the release pipeline, cutting QA effort and recurring production failures in the same quarter.
Mar 2025 — June 2025
NiovLabs
Built core 0-to-1 prototype demos that directly contributed to securing Y Combinator funding, then led the offline evaluation and A/B testing program that followed.
Oct 2022 — Mar 2025
Designing Systems LLC
Took multi-modal AI systems from POC to production behind automated validation gates, then applied TensorRT quantization and ONNX runtime optimization to make inference fast enough — and cheap enough — to run at scale.
Feb 2021 — Aug 2022
Barclays Investment Bank
Built anomaly detection pipelines watching 10M+ daily financial transactions — where I first learned that a model's confidence is only as good as its evaluation harness.
Graduate
Northeastern University
Dean's Graduate Research Assistant
Undergraduate
Yeshwantrao Chavan College of Engineering
Published research with ISRO & NEERI — see Research
Recent research
2026
Yukta Chikate · Zenodo
A risk-adaptive verification framework for consequential AI systems that treats model outputs as candidate actions and introduces runtime verification, evidence provenance, perturbation testing, and ACT / ABSTAIN / ESCALATE decision gates.
2026
Yukta Chikate · Zenodo
A pilot-centered electronic flight bag research prototype studying whether familiar consumer-map interaction patterns can reduce interface friction while preserving aviation-specific state, provenance, and inspectable decision support.
Preprint
Yukta Chikate · arXiv
A runtime assurance architecture for AI systems that separates generation from execution and adaptively allocates verification effort according to consequence, uncertainty, provenance, contradiction, and verification cost.
Public arXiv link not yet availablePreprint
Yukta Chikate · arXiv
An HCI and aviation systems study on selectively transferring familiar consumer-map interaction patterns into pilot-facing electronic flight bags while preserving aviation-specific state, provenance, and decision traceability.
Public arXiv link not yet availableEarlier publications
2024
Northeastern University
Chikate, Y., Brown, N.
2022
Springer · LNNS, Vol. 355
Chikate, Y., Konge, A., Sharma, A.
DOI: 10.1007/978-981-16-8512-5_55 ↗2021
IJERT · ICACT 2021 · Vol. 9, Issue 8
Chikate, Y., Konge, A., Damahe, L.
DOI: 10.17577/IJERTCONV9IS08014 ↗2020
Yeshwantrao Chavan College of Engineering (YCCE)
Chinchole, A., Chikate, Y.
FAA Student Pilot License
Federal Aviation Administration · Issued Aug 2024
General Mobile Radio Service (GMRS) License — Part 95
Federal Communications Commission · Issued Jul 2025 · Expires Aug 2035
Credential ID: FCC call sign — WSJE727
Medical Imaging AI with MONAI Bootcamp
NVIDIA · Issued Nov 2024
2022 – 2024
College of Engineering, Northeastern University, Boston
Research focus: Parkinson's disease analysis, healthcare
2022 – 2024
Healthcare Department, Northeastern University
2019 – 2022
Member — recurring public speaking and leadership practice
2020
Toastmasters International
March 2024
Mentored Yukta
Associate Teaching Professor, Northeastern University
Praised for exceptional technical aptitude and real-world problem-solving across three university courses.
LinkedIn ↗August 2022
Managed Yukta directly
Vice President, Barclays Technology Center India
Instrumental on the team — quickly mastered the FinCrime domain and took on complex challenges without hesitation.
LinkedIn ↗November 2021
Mentored Yukta
Technology Manager, William O'Neil India
A hard-working, fast-adapting engineer who ramped quickly on React.js, Redux, and other new technologies as a fresher.
LinkedIn ↗December 2020
Worked with Yukta on the same team
Principal Scientist, CSIR-NEERI
Delivered detailed GLDAS computation and ArcGIS analysis during a 6-month research internship at NEERI.
LinkedIn ↗