🚫 Please apply only if you are the engineer who will personally do the work. We will not consider agencies, consulting firms, account managers, or subcontracted teams. We're a small growth equity firm based in Connecticut. Our team is using more AI tools internally, and we now need someone senior to help us build the secure Azure infrastructure underneath them. This role is more infrastructure, network, and security focused than AI development. We need someone who is very strong with Azure networking, Entra ID, Key Vault, private endpoints, VNets, firewalls, RBAC, managed identities, and secure access between internal systems. You should also be comfortable with DevSecOps, GitHub-based deployments, secrets management, logging, monitoring, and keeping production environments locked down without making them painful to use. The AI side matters because this environment will run LLM applications, agents, MCP servers, model APIs, and possibly some private or local models. We don't need someone to build prompts or chatbots. We need someone who understands how to securely host and operate AI infrastructure in production. Our systems connect to Microsoft 365, SharePoint, Outlook, CRM data, internal databases, and other business tools, and we handle confidential deal information. We care a lot about network isolation, identity, least-privilege access, auditability, and making sure one compromised service or agent cannot access everything else. This would start around 10-20 hours per week and could grow into an ongoing role. We're looking for an individual engineer, not an agency, and need reasonable overlap with Eastern Time. We'll likely start with a small paid trial around one part of the environment.

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)

We are launching an AI workflow consulting service for small and midsize businesses, initially focused on financial advisory firms, fintech companies and other professional-services businesses. We are looking for a smart, practical AI Workflow Consultant who can help us understand how a client currently operates, identify processes that could be improved and build simple, reliable AI-enabled workflows. This is not primarily a software-development role. We need someone who is excellent at understanding business processes, simplifying them and translating them into workflows using tools such as ChatGPT, Claude, Zapier, Make and other no-code or low-code platforms. Our founder will source clients, lead the sales process and run the initial conversations. You will participate in client assessment calls and take the lead on converting what we learn into an actionable workflow roadmap and working solutions. What You Will Do Participate in 15-minute introductory AI assessment calls with prospective clients Join deeper paid workflow-assessment sessions Ask thoughtful questions about how work is currently completed Map existing processes, including steps, systems, decisions and handoffs Identify repetitive work, bottlenecks and opportunities for AI assistance Rank opportunities based on potential impact and implementation effort Recommend practical tools and workflows Simplify inefficient processes before attempting to automate them Build reusable prompts, custom GPTs, Claude Projects, Claude Skills or no-code workflows Test workflows and refine them based on client feedback Create clear standard operating procedures and training materials Help train client employees to use the finished workflows Coordinate with our developer when a project requires custom technical work Work with our client-service manager to keep projects on schedule Support one active workflow improvement at a time for each client Examples of Potential Workflows Turning meeting notes into summaries, action items and follow-up drafts Preparing client-meeting briefs using approved source materials Repurposing newsletters into emails, social posts and video scripts Creating first drafts of recurring client reports Organizing prospect or client onboarding information Classifying and scoring marketing leads Creating internal knowledge assistants using approved company materials Converting calls or voice notes into structured content drafts Drafting CRM notes and recommended follow-up tasks Documenting repeatable internal processes What We Are Looking For Strong practical knowledge of ChatGPT and/or Claude Experience building AI-assisted business workflows Strong process-mapping and problem-solving skills Experience with no-code tools such as Zapier, Make or similar platforms Ability to understand how a process works before recommending technology Excellent written communication and documentation skills Comfort participating in client-facing calls Ability to explain AI workflows to nontechnical users Good judgment regarding privacy, sensitive information and human review Ability to work independently and deliver against defined acceptance criteria A bias toward simple, maintainable solutions instead of unnecessary technical complexity Experience with financial services, RIAs, fintech, marketing agencies or regulated businesses is helpful but not required. What This Role Is Not It is not a traditional software-engineering position. It is not a role for someone who only provides high-level AI strategy. It is not about building complicated automations when a simple solution will work. It is not an hourly advisory engagement with undefined deliverables. We need someone who can diagnose a process, build the solution, test it, document it and help the client adopt it. Engagement Structure This will begin as a project-based contractor position. We expect to approach three to five existing clients and initially close a small number of pilot engagements. Compensation will be based on completed deliverables rather than hours worked. Before each project begins, we will agree on: The workflow being addressed The expected deliverable The acceptance criteria The fixed payment The anticipated delivery date Initial projects may include: Participating in an AI Workflow Assessment and helping produce the client’s prioritized roadmap Building and documenting one defined AI workflow Supporting the client through testing, training and adoption Continuing with additional workflows through a recurring client engagement If the pilot is successful, this could grow into consistent ongoing work across multiple client accounts.

SANDBULL AI helps owners of trades and local service companies install the AI and automations that already work in their shop. We need a US-based contractor we can pull onto specific client projects as they close. Not a retainer seat. Not a full-time hire. When a project is live, you jump on it, ship it, and we close that ticket. What you will actually do Turn an audit into a build: follow-up sequences, booking, dispatch or estimate flow, inbox, CRM, simple internal tools Connect the stack the client already has (often GoHighLevel, plus whatever they run for jobs, phone, and payments) -Write the SOPs so the owner and their people can run it without us -Join short client calls when we need the implementer in the room -Flag what is out of scope before you start clicking This is not -A sales role -A “prompt engineer” or ChatGPT-wrapper gig -Agency account management - Building our marketing site Must be true -You live and work in the United States (we will only hire US) -You have shipped AI or automation inside real service businesses, not just SaaS demos -You can work from a written brief and a Loom, and you do not need daily standups -You can write clearly to a 50-year-old shop owner -You are fine with project billing (fixed per project or a capped hourly block, we decide per job) Nice if true GoHighLevel, Zapier or Make, and one of: AccuLynx, ServiceTitan, JobNimbus, Housecall Pro, or similar You have sat on a call with a contractor or home-services owner before How to apply Do not send a generic cover letter. In your proposal, take a loom/video and answer these three: 1. One AI project you actually shipped (what you installed, what changed) 2. The tools you used 3.Your hourly rate and a ballpark for a 2-week implementation 4. A goal you are currently working towards in your life. If those are missing, we will skip the proposal.

We are seeking a Forward Deployed Engineer to own full client engagements from pre-sale with a hot lead through delivery and handover. This role combines 60% software engineering with 40% strategic ownership. You will design solutions, build and maintain CI/CD pipelines, assess technical risk, and manage production incidents. You should have 7+ years of experience shipping real-world software. We are tech-agnostic, meaning we are open to considering candidates with diverse technology backgrounds — .NET/C#, Java, Python, Node.js, or similar. While .NET/C# is our primary backend language and Angular our primary frontend framework, prior experience with either is not mandatory — Vue, React, or similar works just fine., with strong backend proficiency in .NET (C#), and experience with cloud platforms. You must also be proficient in production AI integrations, including RAG systems, LLM APIs, agents, and vector databases. You will communicate directly with clients, so having experience in direct client communication is a must; also run requirement discovery sessions, and translate technical constraints into business value without a PM or BA intermediary.

You will work alongside our engineering team inside live customer environments, turning your operational expertise into working AI-powered automations. You have spent years learning how a business actually runs. That knowledge is about to become one of the most valuable inputs in enterprise software, and almost nobody is putting it to work. We are looking for experts from multiple domains who want free training on AI and to solve real world operational problems. What you'll do - Sit with a customer's team and map how a process genuinely operates today - Identify which parts of that process are automatable, which are not, and where the real failure points are - Work directly with Velanir engineers to design, configure, and validate AI agent workflows against that process - Test agent behavior against the edge cases you know will break it, because you have seen them break before - Troubleshoot in production and refine the deployment until it survives contact with reality - Help the customer's team understand and trust what has been built Before you Deploy: Free Training and Certification - Every Forward Deployed Expert (FDEx) completes our 6-week certification program first. It is built and taught by our team of AI experts, and it is free to you. - You will learn how AI agents actually work under the hood, how to scope an automation opportunity, how to design agent workflows, and how to validate and troubleshoot them in production. Not prompt tips. Not theory. The real mechanics of getting agents into production and keeping them there. - You finish with a Velanir FDEx certification and, more importantly, with deployment experience that very few people on earth currently have.

Python Developer Needed – Claude API, PDF Processing & AI Automation I’m looking for an experienced Python developer to help me complete and troubleshoot an AI document-processing workflow. I have approximately 20,000 pages of OCR’d medical records that need to be processed using the Claude (Anthropic) API. The goal is to identify specific mental health evidence while ensuring every extracted finding includes the exact PDF filename and page number so it can be verified and used in court. Current Status * Claude API account is already set up and paid for. * PDFs are already OCR’d and page numbered. * I have an existing Python script that uploads documents and processes them, but it needs improvement. What I Need * Debug and improve the existing Python code. * Ensure page numbers and source filenames are always included. * Improve reliability and error handling. * Optimize processing of a large document set. * Help generate a final organized evidence list. Required Skills * Python * Claude (Anthropic) API or other LLM APIs * API integration * PDF/document processing * OCR workflows * Prompt engineering * Debugging and automation Experience working with legal, medical, or other large document collections is a plus. This is an urgent project, and I’m looking for someone who can start immediately. When applying, please include: * Similar projects you’ve completed * Your experience with Python and AI APIs * Your hourly rate * Your availability over the next couple of days I’m in Las Vegas

##SITUATION We are a boutique, family owned SEC registered RIA serving high net worth families. About 15 people. We run Salesforce Financial Services Cloud, Microsoft 365. I already build with Claude daily: custom skills, projects, MCP connectors, and generated deliverables. What I do not have is a governed system. Brand tokens live in three places. Document output quality depends on who prompted it. Nothing is versioned. Nobody else at the firm can produce a client ready artifact without routing it through me. You will design that system, build the first production version of it, and train me to own and extend it after you leave. If your plan requires you to stay forever, you are not the right fit. ##SCOPE 1. Firm wide Claude design system Condense and build a single governed source of brand truth that every Claude output pulls from. - Brand token file (colors, type scale, spacing, logo rules, chart palettes) structured so it can be consumed by document generation, HTML, and Salesforce components - Skill and project architecture: what belongs in a skill, what belongs in a project, what belongs in a shared reference file, and how they call each other - Naming, versioning, and backup conventions - A written standard I can hand to a new hire 2. Training and handoff on schema and structure - Live sessions (recorded) teaching me how to add, update, and retire skills and reference files without breaking downstream output - A maintenance runbook: what to review monthly, what breaks first, how to test a change before it goes firm wide - Debug patterns for when generated output drifts off brand 3. Document generation pipelines Repeatable, prompt driven production of: -Simple one page formatted outputs (client summaries, meeting recaps, one page briefs) -Complex 20 to 30 page client reports with cover pages, section dividers, running headers, table of contents, charts, and disclosure pages -Excel styles and templates (formatting standards, conditional formatting rules, chart theming, reusable model shells) - Fillable PDF and word docs Each pipeline needs a defined input, a defined output file format, and consistent formatting across runs. Show me how you handle page breaks, chart theming, and long document assembly, because that is where most of these builds fall apart. 4. HTML and LWC component system - Branded HTML component library for microsites, forms, and email friendly layouts - Lightning Web Component patterns for Salesforce Financial Services Cloud that inherit the same tokens, respect SLDS, and pass a Salesforce code review - Documented so Claude can generate new components on pattern rather than improvising 5. Account Engagement email template and block library Salesforce Account Engagement (formerly Pardot) is our marketing automation platform. We need a branded, reusable email system inside it. - Master email template set: newsletter, event invitation, client announcement, advisor introduction, drip and nurture, plain text style personal send - A block or region library our team can assemble without touching code: hero, headline plus body, two column, advisor bio card, event details, CTA button, disclosure footer, social footer - Email safe HTML that renders correctly in Outlook desktop, Outlook web, Apple Mail, Gmail, and mobile, including dark mode behavior - Merge field and dynamic content handling, correct unsubscribe and preference center links, and required disclosure blocks -Tokens inherited from the same brand source as every other output, so a color change happens once - A Claude pipeline that turns a draft or a content brief into a populated, ready to load template Tell me which Account Engagement email builder version you have built in and how you handle testing across clients. We have an existing internal process that converts .eml files into Account Engagement ready HTML. You may extend it or replace it, but say which and why. 6. Voice and positioning system for text generation Codified tone rules by scenario, with examples and counterexamples: - Client email replies - Internal updates and status notes -Client proposals and pitch material - Marketing and newsletter copy - Compliance sensitive language and phrases we never use Deliverable is a voice reference plus the prompt structure that enforces it, not a style essay. 7. Deployment and administration on a Claude Team plan Rollout plan across roles: advisors, planning, operations, service Seat structure, org level feature toggles, and connector policy Adoption plan with training material by role Documented guidance on what can and cannot be restricted at the Team tier, and a recommendation on whether Enterprise is justified for us Compliance context We are SEC registered. Anything client facing is subject to books and records retention and the SEC Marketing Rule. Your build needs to account for review before distribution, source citation in generated reports, and clear rules on what client data may enter a prompt. Prior experience in financial services, legal, healthcare, or another regulated industry is a strong plus. Deliverables checklist Brand token source file plus consumption pattern for each output type Skill and project architecture map with naming and version standards Four or more production document pipelines (one page, long form report, Excel template set, and one of your recommendation) HTML component library plus LWC component patterns with sample code Account Engagement master templates plus a reusable block library, render tested across major clients Voice and positioning reference with scenario prompts Team plan rollout plan, permission configuration, and role based training material Recorded training sessions plus a written maintenance runbook A 30 day and 90 day extension roadmap You are a fit if you have Shipped production Claude skills, projects, or agent systems for an organization, not just personal use Built document generation pipelines that produce polished, repeatable, multi page files Strong design system fundamentals: tokens, components, documentation Working knowledge of Salesforce Lightning Web Components and SLDS Production email HTML experience, including Outlook rendering, dark mode, and marketing automation template systems Comfort writing documentation and teaching a technical but non engineer operator Opinions. I want someone who tells me when my structure is wrong ##Bonus points Financial services or RIA experience Salesforce Financial Services Cloud specifically Salesforce Account Engagement or Pardot template development Excel or Word template engineering at a high level of polish Experience administering a Claude Team or Enterprise deployment

We're a creative production studio making product imagery for major furniture retail brands. Our pipeline is AI-assisted end to end: accurate product renders go in, generated lifestyle scenes come out, with automated quality checks in between. The systems work — we need the person who runs them. The engagement: 15–20 hrs/week to start, running quality review and team cadence for a live image-refresh program. This is contract-to-hire by design: if the trial works, it converts to a full-time role leading this workflow — and eventually the same practice for our other clients. What you'll do: Own the quality gate: review generated scenes against product reference renders, call pass/fail, and send makers specific, actionable notes Keep the batch moving: track who's blocked, keep turnaround predictable, escalate only true edge cases Turn recurring quality issues into written standards and reference examples the team learns from Direct fixes on stubborn generations — different conditioning, stronger reference, different tool — so "try again" is never the note Requirements (hard ones): Production experience with AI image workflows: prompt/reference conditioning, seed iteration, upscaling and compositing. You can name failure modes (identity drift, texture mush, invented geometry) and know which lever fixes each You've managed or led creative makers (CGI, retouch, photo/post production) and made specific people measurably better Sharp eye for product fidelity: wood tone, proportions, hardware — you catch what a generation changed Solid Photoshop compositing fluency. Much of this pipeline lives in layered comps — you need to read them, judge them, and demonstrate a fix when you're teaching one Feedback style that coaches rather than takes over. If your instinct is to fix it yourself, this isn't your gig Not required: building pipelines or writing code. We have engineering; you run the creative side. Start: immediately · Trial scope: the current refresh batch, then a conversion conversation. Propose the rate that reflects your experience — we're hiring for caliber, not to a number.

We're looking for a senior full-stack developer with a strong security background to extend and harden a large, mature application. You'll navigate an existing codebase with real history and real constraints, implement security controls, optimize performance, and keep the system scalable and reliable. You should have deep, self-earned coding fundamentals and know how to use AI as a force multiplier. This is not a role for developers who depend on AI to write code they couldn't write themselves. You should be able to reason through a large legacy codebase unaided — your engineering foundation is what separates AI producing noise from AI producing shippable work. About the Role The stack is TypeScript end to end: a Next.js frontend and AWS-backed services. AI tooling is a deliberate part of the workflow — generating implementation plans, stress-testing those plans against our architecture and business requirements, reviewing generated code for correctness, and shipping with confidence. When AI hits the limits of a complex legacy system, and it will, you're the one who knows how to guide it through. What You'll Do Own the full lifecycle of AI-assisted development: draft plans, pressure-test them against real architectural constraints, and validate that generated code is production-worthy Ensure solutions respect existing codebase structure, clean architecture principles, and layered design patterns Validate that all code — generated or hand-written — ships with appropriate unit and end-to-end tests and thoughtful edge case coverage Apply and enforce security best practices aligned to the NIST Special Publication series on federal security controls: access control, audit logging, system integrity, and secure configuration management across frontend and backend Act as a rapid-response resource for user-facing issues — diagnose, prototype, and deploy fixes fast when production is on the line What We're Looking For Five or more years of proven TypeScript development, with a body of work predating widespread AI coding tools — you know the language, not just the prompts Deep proficiency in React and Next.js Strong working knowledge of AWS (Lambda, SQS, DynamoDB, IAM, and similar) The ability to critically read generated code and catch architectural drift, security gaps, and subtle logic errors the model won't flag itself Familiarity with the NIST federal security control framework or comparable standards, and the judgment to translate controls into practical engineering decisions Comfort using AI tools (Claude Code, Copilot, and similar) as a development partner, with the depth to steer them in unfamiliar or complex codebases Strong debugging instincts and the ability to move fast under pressure without cutting corners on security or quality Bachelor's degree in Computer Science or a related field Nice to Have Mobile platform experience (iOS/Android), infrastructure-as-code, or CI/CD pipeline work. Familiarity with FedRAMP, SOC Type II, or other compliance regimes that map to the NIST control catalog.