Vikram (Vik) Sriram

Senior AI Engineer

  • Enterprise Agentic Systems & AI Platforms
  • Regulated Industries
  • Los Angeles, CA (open to relocation to the Bay Area)
  • phone hidden (hidden, sign in to see it)
  • email hidden (hidden, sign in to see it)
  • github.com/engineer6080
  • U.S. permanent resident

Summary

AI engineer with 6+ years shipping production ML and agentic systems. Architected a global medical-diagnostics company's (hidden, sign in to see it) governed, vendor-neutral enterprise AI agent platform: 6 agents across 4 business domains on open protocols (MCP, A2A, AG-UI), gated by a certification harness, RBAC and governance sign-off; 100+ users and 1B+ tokens (interactive and automated workloads) in its first 4 months. Takes AI from use-case intake to validated production in FDA-regulated environments by solving what blocks enterprise adoption: data security, access control, verification, governance and validation. Background runs from embedded firmware and computer vision through classical ML to multi-agent orchestration, on Azure and AWS.

Skills

  • Enterprise AI & AgentsMulti-agent orchestration; MCP/MCP Apps, A2A, AG-UI; Amazon Bedrock AgentCore, Pydantic AI, LangChain, CopilotKit; LLM gateway and usage telemetry; evaluation harnesses, grounding and validation; human-in-the-loop; tool calling and shared state; OAuth propagation and access gating; vLLM local model serving
  • ML & VisionPyTorch, TensorFlow, scikit-learn; CNNs/MobileNet with Core ML conversion for on-device iOS; OpenCV; multivariate time-series forecasting and ensembles; TF-IDF + BERT text classification; multimodal RAG, document parsing/OCR
  • Data & CloudDatabricks (Unity Catalog, Spark, MLflow), Snowflake, Postgres, Microsoft Fabric, Power BI (custom GenAI visuals); Azure (App Services, APIM, VNet, Private Endpoints, DNS, Entra ID OAuth/SSO/RBAC); AWS (Bedrock, ECS, S3, CloudFormation, PrivateLink); Docker, Kubernetes, Jenkins/Bitbucket CI/CD, OpenTelemetry
  • Embedded & FirmwareC/C++ firmware; Jetson Nano, Raspberry Pi, STM32/ESP32/Arduino; CUDA; UART/USB serial and custom packet protocols; IMU/GNSS drivers; sensor fusion; RTK GNSS, LiDAR, multispectral/hyperspectral sensing
  • Enterprise SystemsSalesforce/ServiceMax and Oracle ERP integrations, Copilot Studio, Power Platform, SmartSolve eQMS
  • Regulated & GovernanceAI governance and use-case intake; non-product software validation (21 CFR Part 820 / QMSR, ISO 13485); risk assessment, Technical Review Board (TRB), CAPA, design control/DFMEA, IP/PII handling
  • LanguagesPython, SQL, TypeScript, C/C++, Swift

Global medical-diagnostics company (hidden, sign in to see it)

Remote (hidden, sign in to see it)Jun 2022 – Present

  1. Senior AI/Data ScientistAI Council MemberOct 2025 – Present
  2. AI/Data ScientistOct 2022 – Sep 2025
  3. Data Science Intern, Advanced AnalyticsJun 2022 – Oct 2022

Enterprise AI Platform & Governance

  • Architect and platform lead for the enterprise AI Agent Catalog, approved by the CTO out of a company-wide continuous-improvement (Kaizen) event (hidden, sign in to see it): 6 governed agents across manufacturing, quality, commercial and finance (OBQ, NC/scrap, manufacturing (hidden, sign in to see it) floor support, cross-domain investigator, IFU search, supplier PO reconciliation) on open protocols (MCP, A2A, AG-UI), 2 in validated production; 100+ users and 1B+ tokens since the June 2026 launch.
  • Built the catalog's certification harness: 50+ tool-level I/O tests per agent plus protocol-level qualitative evaluations; agents earn a production badge only on 100% pass, RBAC enforcement, documentation and TRB / AI-governance sign-off.
  • Established the enterprise LLM gateway, unifying model serving across providers with secure allocation, spend monitoring and per-team usage oversight; model-hosting intake point for company AI intake and platform committee reviews (hidden, sign in to see it).
  • Defined the agent intake, risk-assessment, Technical Review Board (TRB) and governance pipeline with Global Process Owners (GPOs) and cloud ops; wrote onboarding guidance so function teams inherit platform capabilities, integrations and SSO-propagated data-layer security.
  • AI engineer on the enterprise AI Council (operating-company-wide AI strategy planning committee); presented the platform and its governance model to function leaders and coached business teams to build and ship their own GenAI applications.

Production Agents & Workflow Automation

  • Supplier order reconciliation: extracts and matches multilingual supplier documents (~50 vendors) against open Oracle POs using vendor-specific document rules, partial-match rescoring and priority scoring; raised the open-PO line match rate from ~25% (manual) to ~80%, matched on receipt and surfaced in Power BI; 15K+ documents processed; live at a European site (hidden, sign in to see it), regional rollout planned.
  • Instructions for Use (IFU) search: grounded search over a 15K-document controlled corpus for sales, RFP and support teams; 98% accuracy in phase-1 validation vs. ~30% for off-the-shelf AI on complex technical questions, ~80% user thumbs-up; business-estimated ~21.8K hrs/year across sales and technical reps; phase 2 rolling out to 25 sales reps.
  • Out-of-Box Quality (OBQ) daily management: automated a quality team's daily work-order analysis, KPI monitoring and field-service follow-up; ~18–30 hrs/week saved per site, expanding to 2 more sites.
  • Internal manufacturing-floor agent (hidden, sign in to see it): validated production agent for assembly support, troubleshooting and diagnostics.
  • Oracle AR error resolution: took a winning hackathon prototype to production in 3 regions; triages interface errors and proposes fixes for human approval, cutting blocked-sales-order review time ~90%.

GenAI for Quality & Compliance

  • Built GenAI complaint trending across 2.5M+ global complaints, producing traceable month-over-month insights for ~15 SMEs; called "transformative" by the Complaint Handling Unit.
  • Trained and deployed a TF-IDF + BERT ticket-routing classifier (TensorFlow) to production (group/department/subgroup routing); later superseded by an LLM rules engine.
  • Prototyped an LLM-assisted DFMEA engine (hand-built tool routing and state over LangChain, pre-MCP) that cross-references historical DFMEAs, regulatory citations and complaint trends to draft failure-mode entries with full source lineage.

Computer Vision in a Validated Environment

  • Created a vision-inspection app (hidden, sign in to see it), the company's (hidden, sign in to see it) first iOS inspection app to complete full non-product software validation: a marker-free landmark technique uses consistent instrument features in place of ArUco markers, warps a color-coded class template onto the chassis and classifies 32 regions of interest in real time, fully on-device; owned development, enterprise distribution and validation; presented at a medical extended-reality summit (2024) (hidden, sign in to see it).
  • Built glass-diffuser inspection (OpenCV; notch angle and light diffusion, IMU-aligned) and a live supervisor dashboard (2022–23) that contained defect escapes tied to 2 CAPAs and 2 stop-shipments.

Platform & Enablement

  • Stood up and own the AI team's Databricks platform, the first in the org deployed inside the corporate VNet (hub-spoke networking, private subnets, DNS forwarding, VPN-protected Oracle egress), with Unity Catalog, MLflow tracking, model serving and standard notebooks; ~50 users across 2 workspaces.
  • Won 2 of 3 internal hackathons entered (one awarded $100K internal funding for real-time GenAI compliance feedback inside ServiceMax); 2 entries reached production, including a custom GenAI Power BI visual that embeds catalog agents over A2A/AG-UI via an MCP endpoint behind APIM; now program advisor/mentor and lead of org-wide GenAI developer workshops.

Graduate Research Assistant, Computer Science

Cal Poly Pomona

Sep 2020 – Jun 2022

Precision-agriculture robotics

  • Built the autonomy stack for a SuperDroid Robots HD2-S Mastiff tracked UGV for funded precision-agriculture research: reverse-engineered the platform's proprietary drive-control packet protocol, rewrote the Xsens MTi-G-700 GNSS/INS driver for Jetson, ran RTK GNSS waypoint navigation and onboard classification on Jetson Nano with sensor fusion and video streaming on Raspberry Pi 4; 6–7 hour field runs on farm terrain without Wi-Fi.
  • Coordinated DJI Matrice UAVs (MicaSense Altum, Headwall hyperspectral, Applanix GNSS/INS, Phoenix LiDAR) with the UGV: processed terabytes of imagery into radiometrically corrected orthomosaics and 3D point clouds, deployed real-time segmentation on-site, and served a live orthorectified tile map from the UGV stream.
  • Owned field-to-HPC data transmission for the Cal Poly / NASA / USRA fire-detection collaboration; research profiled by CSU STEM-NET (hidden, sign in to see it) (Nov 2020).

Machine Learning Engineer (Computer Vision lead)

Sinsat, Inc. (now Katsh Digital ID)

Los AngelesNov 2019 – Jan 2021

  • Developed hand (dorsal) biometric feature extraction (OpenCV + CNN) and an on-iPhone segmentation engine with offline real-time marker tracking for anonymous identity verification; built the AWS backend and identification API; shipped the first iOS app and website for an 8-person funded proof-of-concept.

Publications, Presentations & Press

Proc. SPIE 12114, Autonomous Air and Ground Sensing Systems for Agricultural Optimization and Phenotyping VII (June 2022). Four peer-reviewed papers, first author on three.

  • Collaboration between UAVs and UGVs for weed classification and detection (hidden, sign in to see it)First author
  • Machine-learning models for the real-time assessment of plant health using UAVs and RGB images (hidden, sign in to see it)First author
  • Improved correlation between UAV-based remote sensing and proximal sensor data (hidden, sign in to see it)First author
  • Estimating health of hybrid bermudagrass managed by soil moisture sensors using remote sensing technology (hidden, sign in to see it)Co-author
  • Medical extended-reality summit poster (2024) (hidden, sign in to see it)Poster

Education & Certifications

  • B.S. Computer Engineering, UC Davis (2018)
  • Graduate studies, M.S. Computer Science program (thesis-track research), Cal Poly Pomona (2020–22)
  • Machine Learning Engineer Nanodegree, Udacity (2019)