We're hiring a senior AI developer to build and deploy AI solutions for a fintech/credit-union platform. The work spans autonomous banking agents, fraud detection, credit scoring, and bill-pay/invoice automation — at the intersection of LLMs, cloud infrastructure, and financial-domain expertise, with security and compliance built in from the start. This is a long-term, ongoing engagement. What you'll do: AI agents & orchestration - Design, build, and deploy multi-agent systems using Amazon Bedrock Agents, LangChain, and related frameworks - Architect agentic workflows for core banking use cases: credit scoring, fraud detection, bill-pay automation, invoice management - Define agent personas, memory strategies, tool-use patterns, and escalation paths for production banking agents LLM engineering - Fine-tune, prompt-engineer, and evaluate LLMs for financial-domain tasks - Build RAG pipelines over credit-union knowledge bases, policy docs, and member data - Implement guardrails, content filtering, and compliance checks for safe, regulated outputs - Monitor performance, hallucination rates, and latency against SLAs Cloud infrastructure (AWS & Azure) - Architect and manage AI/ML workloads on AWS (Bedrock, SageMaker, Lambda, S3, IAM, VPC) and Azure (OpenAI Service, Azure ML, AKS) - Design secure, cost-optimized environments compliant with NCUA, PCI-DSS, and SOC 2 - Implement infrastructure-as-code with Terraform or AWS CDK DevOps & MLOps - Build and maintain CI/CD pipelines (GitHub Actions, Jenkins, CodePipeline, Azure DevOps) - Containerize services with Docker, orchestrate with Kubernetes (EKS/AKS) - Apply MLOps best practices: model versioning, A/B testing, canary deployments, automated rollback - Stand up observability with logging, tracing, and alerting Python development - Write clean, well-tested Python for AI pipelines, REST APIs, and data workflows - Build FastAPI/Flask microservices exposing agent capabilities to frontend and core banking systems - Integrate with financial data sources, core banking APIs, and third-party fintech services Banking applications - Build credit-scoring models using alternative data and explainable AI (XAI) - Develop real-time fraud detection with behavioral analytics, anomaly detection, and auto-decisioning - Create conversational agents for bill pay, account management, and member self-service - Automate invoice workflows: extraction, classification, approval routing, reconciliation - Partner with compliance/risk to keep AI decisions auditable, fair, and regulatory-compliant What you should have: - 5+ years software engineering; 3+ years in AI/ML or LLM engineering - 2+ years building AI for banking, credit unions, or financial services - Hands-on experience with Amazon Bedrock, LangChain, Python, AWS, and infrastructure-as-code - Working knowledge of NCUA, PCI-DSS, SOC 2, GLBA, and Fair Lending requirements - Bachelor's or Master's in Computer Science, Software Engineering, Data Science, or related field Nice to have: - AWS or Azure AI/ML certifications - Open-source LLM experience (Llama, Mistral, Phi) and self-hosted inference (vLLM, Ollama) - Vector databases (Pinecone, OpenSearch, pgvector) - Graph-based fraud networks and graph ML - AI governance / responsible AI framework experience - Prior work at a credit union, community bank, or fintech lending platform To apply, please share: - Your resume highlighting AI and banking project experience - A brief note on your most impactful AI agent or LLM project in a financial-services context - Links to GitHub, portfolio, or published papers (optional but encouraged)

**Overview** We are a fast-growing SaaS company with a lean engineering team (~10 devs) utilizing a modern Python (FastAPI/Django) and Node.js backend, React frontend, and PostgreSQL stack. We have already deployed an initial multi-model agent stack—Claude Code, LiteLLM gateway, Git worktrees, and MCP integrations. We need an expert to run an intensive architecture review and optimization session for our current infrastructure. We are not looking for someone to build a full-time, weeks-long project from scratch. Instead, we need a seasoned engineer who has shipped this exact type of infrastructure end-to-end to audit our setup, identify architectural gaps, and guide our team on hardened implementation. This project must move fast. If your timeline is measured in weeks, please do not apply. We want someone who looks at this scope, jumps into a review session, and delivers actionable architectural guidance in days. This starts as a focused, urgent consultation. However, we expect ongoing advisory work—follow-ups, architecture adjustments, and enhancement reviews—as the AI tooling landscape shifts. For the right engineer, this will turn into a recurring relationship. We are completely open to a fixed price per milestone or an hourly structure. **What You Need to Have Actually Shipped and Can Review (Not Just Read About)** * **Full Agentic Coding Harnesses:** The entire loop: orchestrator → subagent → CI gate → merge loops. * **Isolation Layers:** Configured execution layers (such as E2B, Modal, or secure Docker runtimes) as isolated sandboxes for AI-generated code. * **Parallel Claude Code Sessions:** Managed multiple simultaneous subagents on scoped tasks via Git worktrees. * **Self-Hosted LiteLLM Gateways:** Routing to multiple models (Claude, GPT, Gemini, DeepSeek). * **MCP Server Infrastructure:** Connected file system, PostgreSQL, Atlassian, and Slack tool layers for active agents. * **Agent Framework Structures:** Used CLAUDE.md, COMMON\_MISTAKES.md, subagent role definitions, hook scripts, and settings.json. * **Human-in-the-Loop Orchestration:** Built Plan Mode or equivalent approval gates before agent execution. * **Multi-Agent Frameworks:** 7-agent feature factory patterns or frameworks like LangGraph, CrewAI, or Autogen. * **Durable Workflow Engines:** Applied Temporal, n8n, or similar tools for long-running agent workflow execution. * **Mechanical Quality Gates:** Treating CI green as the ultimate gate for agent output quality. \[[1](https://manveerc.substack.com/p/ai-agent-sandboxing-guide)\] **Our Current Stack (What you are reviewing)** * **Backend:** Python (FastAPI / Django) & Node.js (TypeScript) * **Frontend:** React (Next.js) * **Database & ORM:** PostgreSQL / Prisma / SQLAlchemy * **Infrastructure:** Docker Compose, AWS (ECS/EKS) * **CI/CD:** GitHub Actions / GitLab CI * **AI Layer:** Claude Code with shared `.claude/` directory, CLAUDE.md, and LiteLLM gateway in Docker * **MCP:** Atlassian (Jira/Confluence), custom PostgreSQL MCP server, Slack * **Workflow Automation:** Temporal / n8n * **QA Automation:** Playwright / Autonoma **Scope of Work (Review & Advisory Only)** 1. **Comprehensive Audit:** Audit our current agent harness and identify architectural gaps against a production-grade standard. 2. **Sandbox Strategy Consultation:** Review our environment strategy to ensure highly secure, isolated execution runtimes for agent code runs. 3. **Workflow Hardening Review:** Evaluate our parallel agent workflow setup (Git worktrees, subagent role configs, hook scripts, and settings lockdown). 4. **CI Pipeline Integration Strategy:** Advise on wiring our sandbox execution layer into the existing CI pipeline so agent-executed code runs in clean snapshots, not live infra. 5. **Architectural Runbook:** Deliver an optimization report / documented standard that our backend lead can easily own and execute going forward. **How to Apply** Skip the generic pitch. Show us something real to be considered: 1. A GitHub repo, architecture diagram, or Loom walkthrough of an agentic harness you have actually shipped. 2. Specific tools from our stack you have personally configured (E2B, LiteLLM, Claude Code, etc.). 3. One sentence explaining the hardest problem you solved to get full agent loops running reliably. 4. Your availability to conduct this high-impact architectural review session this week.

Overview I have a Next.js website with a newsletter signup form that currently submits directly from the browser to HubSpot's Forms v3 endpoint. I want to add a lightweight LLM-based spam filter that inspects each submission *before* it reaches HubSpot, and silently rejects (or flags) anything that looks like spam/bot/junk input. Current setup - Framework: Next.js (App Router, TypeScript, React client component) - The form component (`NewsletterForm.tsx`) POSTs directly to `https://api.hsforms.com/submissions/v3/integration/submit/[portalId]/[formGuid]` - Fields collected: `firstname`, `lastname` (optional), `jobtitle`, `email` - Portal ID and Form GUID are public form identifiers (no secrets today) What I want you to build 1. Create a server-side API route in the Next.js app (e.g. `app/api/subscribe/route.ts`) that: - Receives the form fields from the client - Runs an LLM spam/quality check (e.g. OpenAI or similar) to classify the submission as legit vs. spam — checking for gibberish names, fake/disposable emails, nonsense job titles, injection attempts, etc. - If legit → forwards the submission to HubSpot (server-side) - If spam → rejects gracefully with a generic message (no HubSpot write) 2. Update the existing `NewsletterForm.tsx` to POST to the new internal API route instead of calling HubSpot directly. 3. Keep the LLM API key server-side only (use an environment variable — never expose it to the client). 4. Preserve the existing UX: loading / success / error states should still work. Deliverables - Working API route with the LLM spam check + HubSpot forwarding - Updated form component - Brief note on which env vars to set (`OPENAI_API_KEY`, etc.) and how to configure them - Clean, typed TypeScript that matches the existing code style Nice to have (optional) - Basic rate limiting / honeypot field as a cheap first line of defense before the LLM call - Configurable spam threshold or a logged "reason" when something is rejected Requirements to apply - Strong Next.js App Router + TypeScript experience - Experience calling an LLM API (OpenAI or equivalent) from a server route - Familiarity with HubSpot Forms API is a plus To apply, please briefly answer: 1. Which LLM/provider would you use and roughly what would it cost per submission? 2. How would you handle the case where the LLM API is slow or down — do you fail open (let it through) or fail closed (block it)? 3. Have you integrated with HubSpot Forms before? (yes/no is fine)

Overview We're building an open-source CLI gateway for multi-agent AI orchestration — model-agnostic, MCP-native, and designed to bring any agent framework online with a single command. The repo is active, well-documented, and growing. We need an engineer to accelerate integration coverage and help attract open-source contributors. The Work Build agent templates and runnable examples for LangGraph, CrewAI, and similar frameworks Add LLM provider support (Groq, Mistral, Gemini, etc.) to the Hermes runtime Write clean, contributor-friendly code that models good PR hygiene Submit work via fork → PR → merge workflow on GitHub You Are Strong Python developer with CLI tooling experience Familiar with at least one of: LangGraph, CrewAI, LiteLLM, LangChain Comfortable with open source GitHub workflows (fork, PR, issues, reviews) Self-directed — you read docs, ask good questions, and don't wait to be unblocked Nice to Have Experience with MCP (Model Context Protocol) Familiarity with SSE, OAuth 2.1, or agent credential management Prior open source contributions Engagement Part-time to start, 20 hrs/week Fixed milestones per integration delivered Potential to grow with the project To Apply Share your GitHub profile and one example of open source work or a project that shows your Python and agent framework experience. https://github.com/ax-platform/ax-gateway

I need an expert senior software engineer that can provide consulting services around implementation best practices of LLM's and AI into existing application workflows. i.e. leveraging AI to extract data from a document as part of an ingestion pipeline.

Add Tests, Security Audit for Credit Cards, Possible Refactor: Vibe-coded Social Graphify + Lovable: Do not use ai to write your proposal, or for any of your communication. I do not need AI or an AI detection tool to know. It's obscene. I can't trust Claude to make tests. It always slyly reverts when making a productive change to changing the tests to mock data. So I need a human to implement them. Full-stack web dev js, python, llm's. Additional background or interest in: Graph Theory, Graph Neural Nets, Graphical Probabilistic Models, Bayesian Neural Nets, Category Theory, Pre-Deep Learning Natural Language Processing (Cfg's, etc.), Semantic Web Tech (rdf/owl/xbrl), Library Sciences, Pre-LLM Machine Learning (+Stat/Econometrics/etc.), Federated Learning and Crypto is appreciated. Do not use ai to write your proposal, or for any of your communication. I do not need AI or an AI detection tool to know. It's obscene.

We're building an internal AI system that runs entirely on our own hardware (no cloud inference) against our own company data. We have a working proof-of-concept and want to get the architecture right. We need an experienced consultant to review what we've built, pressure-test our decisions, and tell us where we're wrong. This is an advisory/validation role first — we have someone doing the hands-on work; what we want is a senior second opinion to make sure we're building this the right way. What we're running today: Inference: RTX 5090 (32GB, Blackwell), Ubuntu 24.04, running llama-server (llama.cpp + CUDA) serving Gemma 4 31B-it (Q4_K_M GGUF) at a 262,144 context window. Also hosts our MCP retrieval server, PostgreSQL, and Qdrant. Embeddings: separate machine with an RTX 3060 running vLLM serving Qwen3-Embedding-4B. RAG: hybrid retrieval — Postgres full-text search + Qdrant semantic search with RRF fusion, exposed through a custom MCP server with tool-calling. Data: ingesting our own internal operational data into Postgres + Qdrant. Planned stack: LiteLLM for model routing, n8n for automation, Open WebUI for the interface, Langfuse for observability, Vault or Infisical for secrets, Keycloak/Azure AD for SSO. What we need help with: Validating our two-machine split (inference vs. embeddings) and whether our VRAM/context budget holds up under real load — specifically whether a 256K context window is real and performant on a single 32GB card or just nominal. Model selection and routing strategy: which open-weight models for which tasks, and how to structure LiteLLM routes. RAG quality: chunking, embedding dimensionality, hybrid search tuning, reranking — making retrieval actually accurate on messy real-world data. Sanity-checking our overall architecture and telling us our blind spots. You should have done: Stood up local LLM inference in production — llama.cpp/llama-server and vLLM, not just Ollama on a laptop. You understand GGUF quantization (Q4_K_M, IQ-series), KV cache, KV-cache quantization, and how context length maps to actual VRAM consumption. Real fluency in GPU sizing math — given a model, a quant, and a context window, you can tell us whether it fits on a given card and what throughput to expect. Bonus if you've worked with Blackwell / sm_120a. Built production RAG — vector DBs (Qdrant, pgvector), hybrid search, RRF fusion, embedding model selection, reranking, evaluation. Worked with agentic/tool-calling systems and ideally MCP servers. Know the open-weight model landscape (Gemma, Qwen, Llama, Mistral, Phi, Nemotron, Hermes) and their licenses well enough to advise. Production ops: systemd, Docker, model gateways (LiteLLM or similar), observability (Langfuse), secrets management, SSO.

The Role: As a Software Engineer on our AI Infrastructure team, you will help design the core systems that power Prism AI’s generative AI platform. You will help build infrastructure and tools that ensure the reliability, performance, quality, and availability of our AI system. Our mission is to make Prism AI the most reliable and user friendly generative AI platform in the world. You will partner closely with our cloud infrastructure team, product team, and performance team to deliver infrastructure that bridges the gap between our customer and the ultra-performant proprietary Prism inference engine. Key Responsibilities: Contribute to the design and development of scalable backend infrastructure that supports distributed training, inference, and data pipelines Build and maintain core backend services such as LLM CI/CD pipeline, control plane, and model serving systems Support performance optimization, cost efficiency, and reliability improvements across compute, storage, and networking layers Building frameworks and safeguards to ensure Prism AI has the best model quality in the industry Collaborate with performance, training, and product teams to translate research and product needs into infrastructure solutions Participate in code reviews, technical discussions, and continuous integration and deployment processes Minimum Qualifications: Bachelor’s degree in Computer Science, Engineering, or a related technical field (or equivalent practical experience) 3 years of experience in software engineering, with a focus on infrastructure or machine learning systems Strong programming skills in Python, Go, or a similar language Proven experience in ML infrastructure and tooling (e.g., PyTorch, MLflow, Vertex AI, SageMaker, Kubernetes, etc.). Basic understanding of LLM knowledge (e.g., context length, disaggregated prefill, KV cache memory estimation, etc) Preferred Qualifications: 5+ years of experience in software engineering, with a focus on infrastructure or machine learning systems Experience with open source inference engine like vLLM, Sglang, or TRT-LLM Contributions to open-source infrastructure or ML projects Experience in building large scale ML/MLOps infrastructure

NobleProg is seeking an experienced AI Trainer to deliver a live, instructor-led remote training focused on helping technical professionals integrate Agentic AI and RAG systems into their existing workflows. This opportunity is designed for participants with strong technical backgrounds (Data Engineering and Workflow Automation) but limited formal AI experience, with the goal of applying AI to real-world systems rather than learning theory. Engagement Details Location: Remote Duration: 2 days Audience: Data Engineers and Workflow Developers Participants: 4+ Daily Rate $1,100 per day Course Scope This training focuses on practical, hands-on development of AI-powered systems using Retrieval-Augmented Generation (RAG) and agent-based architectures. The course will follow a Core & Split approach, starting with shared foundational concepts, moving into role-specific deep dives, and concluding with an integrated session demonstrating how AI systems are built and applied across workflows and data pipelines. NobleProg SOP - https://share.synthesia.io/a0788c6e-56d5-4da8-92c6-0d5c03ad6d52 Key Topics Include - Practical introduction to LLM applications and AI system architecture - Retrieval-Augmented Generation (RAG) design and implementation - Data preparation, embeddings, and vector database concepts - Agentic AI fundamentals (tools, decision-making, multi-step workflows) - Orchestration frameworks such as LangChain, LangGraph, or similar - Role-based applications: RAG pipelines for data engineers and AI-driven workflows for workflow developers - End-to-end system integration (RAG + agents + automation) Trainer Responsibilities - Deliver engaging, instructor-led remote training with strong hands-on focus - Translate AI concepts into practical applications for non-AI technical professionals - Structure delivery using a Core & Split model to address different roles - Provide real-world exercises aligned with data pipelines and workflow automation - Facilitate an integrated session demonstrating how different components work together - Prepare training materials (trainer retains ownership of content) Required Qualifications - Hands-on experience building LLM-based applications, including RAG systems and agent-based workflows - Strong proficiency in Python and experience with APIs, data pipelines, or automation systems - Experience with frameworks such as LangChain, LangGraph, or similar - Proven experience delivering technical training to engineering audiences - Ability to simplify AI concepts and connect them to real-world use cases Nice to Have - Background in data engineering, workflow automation, or solutions architecture - Familiarity with MCP or emerging agent orchestration frameworks - Experience designing modular or role-based training programs preferred - Experience building production-grade AI applications preferred https://docs.google.com/document/d/184VlJipyixkLNJ_HnP3aPt4YToedTUAlji_LxkuLhRU/edit?usp=sharing Please review and approve this tentative outline. We will be meeting with the client to determine whether they prefer a 1-day or 2-day delivery format. The agenda may require some adjustments based on the client's specific objectives, technical background, and areas of interest, which can be finalized during the trainer-client consultation call. Could you please review the proposed outline and let us know if you see any red flags, gaps, concerns, or topics that may require immediate attention? We would also appreciate any recommendations regarding scope, level of technical depth, hands-on exercises, or prerequisite knowledge that should be addressed before presenting this to the client. Thank you for your feedback. How to Apply Please include - A brief overview of your experience with Agentic AI and RAG systems - Your experience delivering technical or AI-focused training - Examples of AI systems or applications you have built - Your approach to teaching participants without formal AI background - Availability for remote delivery

Eligibility: This role is open to U.S. citizens only due to client security and compliance requirements. Please apply through this posting only — do not contact Data-Sleek directly regarding this position. Applications received outside this channel will not be considered and reported to Upwork. Data-Sleek is looking for a Senior AI Solutions Engineer to lead our on-premise and government-cloud AI deployments. You will design, build, and deploy AI-powered data pipelines for clients who cannot use commercial cloud due to ITAR, CMMC, or other data residency constraints, beginning with a client in the aerospace and defense sector. Beyond this first engagement, you will become Data-Sleek's go-to engineer for AI deployments across defense and aerospace clients, building the practice rather than just executing a single project. About Data-Sleek Founded in 2020, Data‑Sleek® is a U.S.-based AI and data consulting firm that helps mid-market companies build the data foundation that AI actually runs on. We own the full path — data strategy, architecture, integration, warehousing, and AI implementation — so organizations can adopt AI with confidence, stay compliant, and scale, without first hiring an internal data team. Our distributed U.S. team (San Francisco, Los Angeles, Irvine, Dallas, Chicago, and New York) partners with clients across healthcare, finance, insurance, logistics, and technology, modernizing data platforms with best-in-class tools like Snowflake, dbt, Fivetran, Tableau, and AWS. Trusted by Fortune 500 institutions and growing companies alike, Data‑Sleek turns complex data into measurable outcomes — faster insight, lower cost, and AI projects that deliver. About the Role You will own the technical delivery of AI-powered data pipelines in restricted environments where commercial cloud is not an option. The immediate engagement centers on a Product Lifecycle Management (PLM) data migration: building a pipeline that connects to a client's SharePoint on a restricted Microsoft 365 government tenant, reads engineering documents, classifies and summarizes them, detects duplicates, and rates naming-convention compliance to produce a migration-readiness report. You will start on-premise, then help the client evaluate and move to government cloud for production. Key Responsibilities AI Pipeline Development Build AI pipelines that connect to a client's SharePoint on a government cloud tenant, read engineering documents, classify them by type, generate summaries, detect duplicates, and rate naming-convention compliance in support of PLM data migration. Catalog large document repositories and produce migration-readiness reports and Excel catalogs that give clients a clear, measurable picture of their data. Engineer document-parsing workflows across DOCX, PDF, and XLSX formats, including embedding generation and database operations. On-Premise & Government Cloud Deployment Deploy on-premise first — a Mac Mini running Gemma via Ollama — standing up, serving, and tuning local inference infrastructure. Evaluate and migrate to production on Azure OpenAI (Azure Government) or AWS Bedrock (GovCloud) when the client is ready to scale. Keep deployments compliant within ITAR-sensitive, restricted-network boundaries throughout. Architecture & Cost Advisory Produce cost models and architecture recommendations that help client IT teams make informed platform decisions based on measured data, not vendor pitches. Compare deployment options — local, Azure Government, and AWS GovCloud — on cost, performance, and compliance, and explain the trade-offs clearly. Practice Building & Delivery Serve as Data-Sleek's go-to engineer for AI deployments across defense and aerospace clients. Build a reusable capability — a repeatable AI-solutions practice — rather than executing a single one-off project. What You Bring Required U.S. Citizen: U.S. citizenship is required and non-negotiable due to ITAR and client security and compliance requirements. Production LLM deployment: You have stood up inference infrastructure — not just called an API. You've handled model loading, memory constraints, failure modes, and throughput tuning in a real deployment. Local inference: Ollama, vLLM, llama.cpp, LM Studio, or TGI. You've served open-source models (Gemma, Llama, Mistral) on local hardware. Cloud AI platforms: Azure OpenAI or AWS Bedrock — at least one. Service configuration, model access, authentication, and token-based pricing. Python: Pipeline engineering — document parsing (DOCX, PDF, XLSX), API integrations, embedding generation, and database operations (SQLite, Postgres). Experience: 5+ years post-degree in software engineering, data engineering, or ML engineering. Strong Preferences Microsoft ecosystem: Entra ID, Microsoft Graph API, and SharePoint REST API at the API level. GCC High experience is a bonus. MCP (Model Context Protocol): Experience building or consuming MCP servers — a significant plus for a fast-evolving protocol. Workflow orchestration: n8n, Temporal, Airflow, or similar. The pipeline is orchestrated, not scripted. Government cloud awareness: Understanding of what FedRAMP High, IL4/IL5, and ITAR mean for cloud architecture decisions. Embeddings & vector similarity: sentence-transformers, pgvector, Qdrant, or FAISS for duplicate detection. 
Bonus (valued if present) Aerospace or defense experience: Familiarity with ECOs, BOMs, and AS9100 saves ramp time. Apple Silicon optimization: MLX, Metal acceleration, and Ollama tuning on M-series chips. Agentic frameworks: Bedrock AgentCore or Azure AI Foundry — the future direction involves agentic AI workflows on government cloud. What This Role Is Not Model training or fine-tuning. This is deployment engineering, not research. Data science or statistical modeling. The AI here is document understanding and classification, not predictive analytics. Frontend development. The deliverable is an Excel catalog and a report, not a web app. Sales or client acquisition. Data-Sleek's leadership manages the client relationship; you focus on delivery. Engagement & Compensation Remote, US-based. Occasional on-site travel to client facilities for hardware deployment and workshops may be needed. An average of 2–3 trips for the first engagement may be possible. Compensation. $40-$55/hour Why Join Data-Sleek? At Data-Sleek, you'll lead AI deployments in environments most engineers never touch — government cloud and on-premise systems where commercial tools simply aren't an option. Your work will directly shape how defense and aerospace clients adopt AI, and you'll build a reusable capability the company grows around. We focus on doing the right thing architecturally rather than selling the most expensive option, and we give our engineers the autonomy to deliver real solutions for real constraints. How to Apply If you've shipped real LLM deployments with real constraints, we'd like to hear from you. Please submit: Your resume A brief note describing one LLM deployment you've shipped — what model, what infrastructure, what data source, and what went wrong. Data-Sleek® is an equal opportunity employer. We celebrate diversity and are committed to creating an inclusive environment for all contractors.