Hi, I'm Akshay Patel

Founder & Applied AI Engineer building deterministic AI automation systems and agentic workflows.

Applied AI engineer specializing in multi-agent orchestration, enterprise automation, and LLM guardrails - building systems that make autonomous workflows reliable, evaluable, and production-ready.

Google ScholarOSF Research(soon)
5+
Years Building Systems
4+
AI Agent Products
3
Featured AI Systems

Research & Publications

I study how to make agentic systems behave predictably in production: clearer failure modes, stricter tool boundaries, and evaluation that measures reliability - not just demo quality.

Ongoing research
DraftOngoing research · 2026

Deterministic Execution Constraints in Autonomous Multi-Agent Workflows

Problem
In multi-agent pipelines, one bad tool call or fuzzy handoff can cascade: wrong API writes, skipped verification, or loops that never converge. I’ve seen this in brand-compliance agents, support agents, and cross-system automation bridges - where “mostly correct” is not good enough for live workflows.
Approach
I'm documenting recurring failure patterns (tool misuse, state drift, weak verification), then designing explicit constraints: allowed tools per step, typed state transitions, retry budgets, and human/system gates when confidence is low. The goal is a practical constraint set you can apply to LangGraph-style and similar orchestrations - not a vague reliability checklist.
Next
Turn the draft notes into a short technical write-up with examples from current agent builds, then publish a PDF (GitHub or OSF) for feedback before any formal preprint.
Multi-Agent SystemsGuardrailsDeterministic Workflows+1
PDF (soon)DOI (pending)OSF (soon)
Ongoing research
In ProgressOngoing research · 2026

Evaluation Benchmarks for Tool-Using Agents in Enterprise Workflows

Problem
Most agent demos are judged by a few happy-path chats. In enterprise settings you also need latency budgets, grounded tool outcomes, and clear fail/escalate behavior. Without a shared task set, it's hard to tell if a change helped - or just looked better in one transcript.
Approach
I'm defining a compact benchmark: fixed tasks, allowed tools, success criteria, and metrics for latency, hallucination/off-tool actions, and escalation correctness. Early design targets support-style and ops-style workflows (ticket context, multi-system lookups, structured updates) so results map to work I already ship.
Next
Freeze v0 task definitions, run baselines on current calling/chat agents, and record scores so later iterations have something to beat.
EvaluationTool-Using AgentsBenchmarks
PDF (soon)DOI (pending)OSF (soon)
View all research

Draft / In Progress = active research tracks (not published papers yet)

About Me

Founder & Applied AI Engineer focused on deterministic automation, multi-agent workflows, and production-ready LLM systems.

My Journey

I'm Akshay Patel, a Founder and Applied AI Engineer with 5+ years building software systems - now specialized in multi-agent orchestration, enterprise automation, and LLM guardrails. My work sits at the intersection of research-minded design and production delivery.

I design agent pipelines that are evaluable and constrained: LangGraph workflows, RAG and ranking systems, FastAPI services, and verification loops that reduce hallucinated actions. Earlier I shipped full-stack and mobile products at scale; that systems background now informs how I harden AI for real latency, reliability, and tool-integration constraints.

Outside shipping systems, I write technical reports, mentor on generative AI, and explore AI safety evaluation - work aimed at fellowships and research collaboration as much as product impact.

Core Focus Areas

Multi-Agent Systems

LangGraph, orchestration, tool-use guardrails

Applied LLMs

OpenAI/Claude APIs, RAG, evaluation loops

Production AI Backends

FastAPI, PostgreSQL, Redis, Docker

Reliability & Guardrails

Hallucination mitigation, deterministic constraints

Cloud & Delivery

CI/CD, containerized AI services

Technical Leadership

Mentoring, research labs, cross-functional delivery

Akshay Patel - Professional Photo

Key Achievements

5+ Years Building Systems

From production apps to agentic AI pipelines

Applied AI Research Focus

Deterministic agents, evaluation, guardrails

Multi-Agent Products

Brand Craft, advising systems, GovGuide agent

Enterprise Automation

Latency-aware workflows and tool integrations

Interested in deterministic agents, evaluation, or research collaboration?

Skills & Expertise

Technical stack organized for applied AI systems work: multi-agent orchestration, evaluation, and production infrastructure.

Core Skills & Tech Stack

Tier 1

LLMs & Multi-Agent Systems

Agent orchestration, model APIs, and deterministic workflow control

LangGraph90%
AutoGen85%
OpenAI / Claude APIs92%
Multi-Agent Orchestration90%
LLM Guardrails88%
Prompt & Tool Design90%

Tier 2

ML & Evaluation Systems

Retrieval, model training, and reliability measurement

PyTorch82%
RAG Pipelines88%
Vector Databases85%
Evaluation Benchmarks84%
Hallucination Mitigation86%
Embeddings & Ranking83%

Tier 3

Production Systems

APIs, data stores, and deployable services for AI workloads

FastAPI88%
Redis82%
Docker85%
PostgreSQL84%
Python / Node.js90%
CI/CD & Cloud Deploy80%

Soft Skills & Research Practice

Systems Thinking

Designing agent pipelines with clear failure modes, constraints, and evaluation loops

Research Rigor

Turning applied experiments into reproducible benchmarks and technical write-ups

Technical Leadership

Mentoring engineers and aligning multi-agent builds with product and safety goals

Problem Decomposition

Breaking ambiguous AI automation goals into deterministic, testable workflows

Cross-Functional Delivery

Shipping with research, product, and ops partners under real latency and reliability constraints

Reliability Focus

Prioritizing guardrails, observability, and hallucination mitigation in production agents

Research Focus & Learning

AI Safety Evaluation

In Progress

Deterministic Agent Control

Researching

Multi-Agent Benchmarks

Exploring

Systems for LLM Ops

Learning

Ready to apply these systems in research or production?

Featured Projects

High-impact applied AI systems: multi-agent orchestration, recommendation pipelines, and domain-grounded agents. Full catalog lives on the projects page.

StickLab
Completed

StickLab

AI sticker studio with vision segmentation, multimodal generation, and automated design workflows.

Vision APIsImage SegmentationMultimodal AIAutomated Design PipelinesPythonComputer Vision
AI Academic Advisor
Completed

AI Academic Advisor

Hybrid recs + NLP/LSTM engagement models with latency-tuned Flask APIs and explainable advising.

PythonFlaskReactTypeScriptPostgreSQLMongoDBTransformersLSTMNLP
GovGuide Florida: AI Agent
Completed

GovGuide Florida: AI Agent

Domain-grounded multi-step agent: search → scrape → synthesize for Florida government Q&A.

Relevance AIAI AgentsLLMsPythonGoogle Search APIWeb ScrapingJSON

Showing 3 featured AI systems projects. Browse the full catalog and earlier works next.

View all projects

Experience & Education

Professional path across applied AI systems, generative AI teaching, and production engineering - from multi-agent deployments to research-minded evaluation practice.

IT Software & System Engineer

Allied Digital Services
Jan 2026 - Present
United States

Building applied AI automation inside an enterprise services environment: custom AI agents, voice/calling agents, and chat agents that handle operational workflows, plus AI integration bridges that connect disparate internal and customer software systems. Focused on deterministic tool use, multi-system orchestration, and reducing manual handoffs across support and business processes.

Technologies & Skills

PythonLLM APIs (OpenAI / Claude)LangChain / LangGraphAI Voice & Calling AgentsChat AgentsMulti-Agent OrchestrationAPI Integration BridgesWebhook / Event WorkflowsPrompt EngineeringServiceNow / Enterprise Systems APIs

Key Achievements

  • Designed and deployed customized AI agents for internal and client-facing operational workflows beyond traditional desk-side support
  • Built AI calling (voice) agents for outbound/inbound process automation and structured call handling
  • Implemented chat agents that resolve routine requests and route complex cases with tool-backed responses
  • Created AI bridges that connect heterogeneous software systems (tickets, CRM/ops tools, and internal services) through APIs and orchestrated agent tools
  • Reduced manual cross-system handoffs by automating data lookup, status updates, and multi-step task flows with guardrailed agent actions
  • Iterated prompt, tool, and evaluation loops to improve reliability of agent outputs in live business workflows

Software Engineer Intern

MSL Techsolution
Jul 2025 - Dec 2025
Tallahassee, Florida, United States (Remote)

Designing and deploying end-to-end AI agent solutions for enterprise clients, focusing on lead generation, customer support, and marketing automation. Engineering scalable, multi-agent systems and custom generative AI pipelines using LangChain, Transformers, and various LLM APIs.

Technologies & Skills

PythonLangChainOpenAI APIHugging Face TransformersDALL-E 3Multi-Agent SystemsPrompt EngineeringDockerKubernetesCI/CDGoogle CloudNo-Code AI Tools

Key Achievements

  • Designed and deployed LangChain-powered voice and chat agents, reducing manual client outreach by 40%
  • Improved client call conversion rates by 25% by developing a lead generation voice agent trained with advanced prompt engineering and Hugging Face Transformers
  • Engineered a custom AI image generation pipeline using the OpenAI API to produce brand-specific creative assets, accelerating client marketing campaigns
  • Built and managed Kubernetes-orchestrated deployments on Google Cloud, ensuring scalability and automated CI/CD pipelines for AI microservices
  • Reduced prototyping experimentation cycle time from days to hours by researching and implementing no-code AI agent flows

Teaching Assistant

San Francisco Bay University
Feb 2024 - Jul 2024
Fremont, California, United States

Mentored graduate students in advanced Generative AI topics, focusing on first-principles understanding of LLM architectures, embeddings, and agent-based systems. Facilitated hands-on learning through labs, projects, and tool-based experimentation.

Technologies & Skills

Database SystemsLangChainOpenAI APITransformersHuggingFaceAdvanced Web DevelopmentPrompt EngineeringPythonFlaskLLM Agents

Key Achievements

  • Guided 20+ students on building and deploying LangChain-based AI agents with custom memory modules
  • Designed assignments covering embedding models, prompt engineering strategies, and transformer tuning
  • Conducted live labs on multimodal pipelines, agent orchestration, and conversational AI systems
  • Reviewed student projects and optimized LLM outputs through systematic evaluation and feedback loops
  • Contributed to curriculum improvement by integrating real-world use cases and toolchain updates

Software Developer

e-infochips Private Limited
Nov 2022 - Aug 2023
Ahmedabad, Gujarat, India (Remote)

Engineered AI-powered dashboards, real-time mobile apps, and scalable microservices across cloud and IoT systems using a modern full-stack toolchain. Collaborated on data-driven UX and GenAI integrations for predictive, accessible, and high-uptime enterprise solutions.

Technologies & Skills

TypeScriptReact NativePostgreSQLAWS LambdaDockerGraphQLJestGitHub ActionsLangChainOpenAI APIFirebase

Key Achievements

  • Improved system performance by 40% by optimizing mobile dashboards and UI rendering logic
  • Engineered real-time data pipelines and Lambda-based services to support IoT telemetry processing and predictive alerts
  • Integrated REST and GraphQL APIs, reducing API response time by 25% and enhancing cross-device responsiveness
  • Built reusable CLI scaffolds using Cookiecutter, reducing app setup time by 70%
  • Led logging utilities and UI component standardization, cutting incident debugging time by 55%
  • Resolved over 50% of critical production bugs within 24 hours, achieving 99.9% uptime
  • Implemented CI/CD pipelines via GitHub Actions and Docker, accelerating deployment cycles by 35%
  • Collaborated with product and design teams to deliver monitoring dashboards used across 3+ internal teams

Software Developer

Moon Technolabs Pvt. Ltd.
Feb 2019 - Nov 2022
Ahmedabad, Gujarat, India

Led the development of real-time cross-platform applications using React Native, WebSockets, and Firebase. Delivered performant, accessible, and scalable mobile frontend and backend integrated with secure backend APIs for global clients across logistics, finance, and travel sectors.

Technologies & Skills

React NativeReduxFirebaseWebSocketsAWS S3Node.jsDjangoGoogle Maps APIGraphQLJestWebGLAWS LambdaDockerGitHub ActionsPostgreSQL

Key Achievements

  • Delivered 10+ production-ready apps with full-cycle deployments to App Store and Play Store
  • Integrated REST and GraphQL APIs, reducing API response time by 25% and enhancing cross-device responsiveness
  • Built OCR-based form auto-fill logic using AWS S3 and image parsing, cutting manual input time by 60%
  • Built reusable CLI scaffolds using Cookiecutter, reducing app setup time by 70%
  • Led logging utilities and UI component standardization, cutting incident debugging time by 55%
  • Resolved over 50% of critical production bugs within 24 hours, achieving 99.9% uptime
  • Implemented CI/CD pipelines via GitHub Actions and Docker, accelerating deployment cycles by 35%
  • Collaborated with product and design teams to deliver monitoring dashboards used across 3+ internal teams
  • Improved app performance by 30% using real-time messaging, lazy loading, and optimized state management
  • Integrated WebGL/Godot-powered 3D UI components, boosting engagement in specialized use cases
  • Led UI/UX optimization efforts to ensure pixel-perfect, accessible, and cross-device compatible layouts
  • Mentored 4 junior developers and enforced standardized code practices through design/code reviews
  • Applied performance strategies including bundle splitting and memoization, reducing load time by 35%
  • Contributed to Agile sprints and delivered 3+ client-facing products with secure, responsive interfaces

Python Developer Intern

Virtual Heights Pvt. Ltd.
Jul 2018 - Feb 2019
Ahmedabad, Gujarat, India

Developed a real-time IoT luggage tracking system using Django and Raspberry Pi, focusing on biometric security, dashboard visualization, and synchronized multi-platform data handling for physical asset protection.

Technologies & Skills

PythonDjangoRaspberry PiMySQLREST APIsBootstrapFlask

Key Achievements

  • Engineered a fingerprint-secured IoT tracking system integrated with Raspberry Pi, boosting physical security by 70%
  • Built a responsive Django admin dashboard for real-time location tracking, alerting, and device visibility
  • Created secure RESTful APIs to synchronize data across mobile, web, and edge devices
  • Implemented encrypted communication, live status updates, and fault-tolerant data sync logic
  • Contributed to early-stage security design and session logging for audit compliance and device diagnostics

Open to research collaboration, AI systems roles, and fellowship conversations.

Let's Connect

Based in the United States. Open to research collaboration, AI systems roles, fellowship conversations, and applied multi-agent projects.

Get In Touch

Based in the United States - open to research collaboration, AI systems roles, fellowship inquiries, and applied multi-agent / automation projects.

Location

United States

Follow Me

Send a Message

Fill out the form below and I'll get back to you as soon as possible.

Interested in working together?

Whether it's agentic workflows, evaluation systems, research collaboration, or production AI automation - let's talk.

Send Email

Akshay Patel - Founder & Applied AI Engineer

Founder and Applied AI Engineer building deterministic AI automation systems, multi-agent orchestration, and LLM guardrails for production workflows.

Expertise includes: Multi-Agent Systems, LangGraph, AutoGen, RAG Pipelines, Evaluation Benchmarks, FastAPI, Docker, PostgreSQL, Applied ML Systems Research.

Available for research collaboration, AI systems work, and fellowship inquiries.