Portfolio

AI systems that drive real impact

We build AI products, internal agents, and company-brain systems for real operating work across client services, fintech, GTM, engineering, platform teams, and select studio MVPs.

Technologies used across these builds

TypeScript

Railway

Postgres

Saving 40+ hours each week for a financial services operator

Customer enquiries used to move through marketing channels, forms, CRM entry, WhatsApp follow-up, and consultant handoff by hand. Grail connected that intake path into a reviewed agent workflow.

Read case study

Problem

The team was receiving enquiries from TikTok, Meta, website forms, WhatsApp, and other channels. Each lead still had to be cleaned up, entered into the CRM, followed up manually, qualified, and handed to a consultant. Good leads could sit too long before the first response.

Solution

Grail connected the lead sources, CRM, WhatsApp follow-up, qualification questions, and consultant handoff into one reviewed workflow. The agent handles the first response and routing while the human team keeps ownership of advice and sales conversations.

Impact

Reported impact: 40+ hours of manual work removed each week and a 40% revenue lift from faster lead response.

Fintech

Human-reviewed transaction checks for a fintech operator

Sensitive transaction checks needed evidence, reconciliation context, and reviewer control before any writeback. Grail built a workflow that prepares the packet while humans keep the final decision.

Read case study

Problem

Transaction checks and reconciliation were sensitive, repetitive, and hard to automate safely. The team needed help gathering evidence and preparing review work, but not an unconstrained agent making compliance decisions on its own.

Solution

Grail built a review workflow that gathers transaction and AML/KYT evidence, applies clear scoring rules, stores the trail, and stages decisions for human approval before any writeback.

TypeScriptAML/KYT APIsApproval gates

Impact

Reviewers get a consolidated evidence packet and approval step before sensitive reconciliation or compliance actions move toward writeback.

Fintech

Reducing SOP support inside a fintech operator

Operations questions were scattered across SOPs, team messages, and informal handoffs. Grail built a Teams-native agent lane to answer repeat questions and package messy issues for review.

Read case study

Problem

Operations work was happening inside messages, documents, and repeated support questions. The team needed faster answers from SOPs and a cleaner way to turn informal issues into tasks without giving an agent broad write access on day one.

Solution

Grail created a dedicated Teams lane with its own app package, document access path, deployment wrapper, and staged rollout. The first phase focused on read-only Q&A, issue intake, and structured handoff packages.

Impact

The agent reduced repeated SOP support work and made product/vendor handoffs easier to package for review.

Workforce operations

Internal operations platform for a workforce operations team

A workforce operations team needed a narrow internal app for a specific workflow. Grail kept the scope small and shipped the operator surface instead of turning it into a broad platform rollout.

Read case study

Problem

The team needed a practical internal app for a defined operations workflow, not a broad platform rebuild. The useful path was a narrow build that captured the workflow, gave operators a usable surface, and stayed small enough to ship quickly.

Solution

Grail scoped the workflow, built the internal app surface, and handed over a small operator-facing tool for the client team.

Stack used

Internal appWorkflow specOperator UI

Impact

The finished app gave the team a focused workflow surface and showed where smaller implementation projects can create value quickly.

Internal tools & open source

From internal operations to reusable infrastructure

The same implementation patterns behind client work also power Grail's own AI employees, agent runtime, and open-source platform projects.

Internal tool

Outbound and research engine

Outbound research used to lose context across CRM records, websites, LinkedIn evidence, and notes. This workflow keeps the evidence attached and turns it into reviewed outreach queues.

Read case study

Problem

Outbound work was spread across CRM records, websites, LinkedIn, notes, public data, and hand-written drafts. Research evidence was easy to lose, and follow-up quality depended too much on whoever did the research that day.

Solution

Grail built an engine that enriches CRM records, captures evidence, explains why an account may be relevant, and drafts LinkedIn or email outreach for human review.

Impact

The workflow preserves evidence, helps prioritize better-fit prospects, speeds up outreach preparation, and keeps final sending human-reviewed.

Internal tool

Development agents for engineering work

Engineering work often spans repo inspection, code edits, browser checks, and verification. Grail’s development agents carry that loop end to end and report back with the checks they ran.

Read case study

Problem

Engineering tasks often require reading repo rules, editing code, running commands, checking browser behavior, and reporting what changed. A single chat response cannot safely hold that much state or finish the work end to end.

Solution

Grail combined Codex-style workers, repo context files, browser control, command execution, verification loops, and long-running task support into a repeatable engineering workflow.

Stack used

Codex CLI

Rust workersMCPBrowser automationOpenflow

Impact

The same development-agent loop now operates across the Grail workspace and informs platform work like Openflow and Grail AI OS.

Internal tool

Slack AI employees for internal work

Grail teams ask for work where the conversation already happens: Slack. These AI employees read context, use tools, work on files and code, and report back with what changed and how it was checked.

Read case study

Problem

A normal chatbot can answer questions, but it cannot reliably carry multi-step work for a team. Grail needed agents that could work where the team already asks for help, keep context, use tools, and report back clearly.

Solution

Grail built Slack-based AI employees backed by a durable task runtime, Codex-style workers, browser and file tools, memory, approvals, and verification reports.

Impact

The agents now handle repeated internal work across code, research, browser workflows, reports, and team operations.

Open source

AgentOS platform layer

AgentOS is the platform layer behind Grail agents: long-running work, memory, files, tools, approvals, logs, permissions, and self-hosting.

Read case study

Problem

Companies cannot use AI safely in real business processes if the AI only lives in a chat window. They need agents, workflows, records, approvals, permissions, logs, and deployment control in one system.

Solution

Grail built the AgentOS/Grail AI OS direction around business objects, agents, workflows, API/MCP control, LLM routing, sandboxed execution, connectors, and egress controls.

Impact

The repo is a launch-candidate platform scaffold focused on architecture, governance, and self-hosted distribution readiness rather than customer adoption metrics.

Open source

Spacetime workspaces

Spacetime is a persistent compute layer for creating workspaces, running commands, serving apps, checkpointing state, and forking environments.

Read case study

Problem

Agent workloads depend on compute environments that can outlive one command, stay inspectable, resume later, checkpoint before risky changes, and fork from a known-good state.

Solution

Grail built a Rust-first workspace system with microVM runners, copy-on-write volumes, control-plane APIs, gateway routing, admin visibility, and CLIs.

Impact

The local repo includes CLI quickstarts, GCP deployment scripts, health checks, checkpoint and fork commands, and a read-only admin dashboard path.

Open source

FastClaw agent runtime

Channel agents need durable execution, not just chat replies. FastClaw queues work, scopes tools and context, handles memory and approvals, and returns results to the right channel.

Read case study

Problem

Teams want agents inside Slack, Teams, WhatsApp, and other channels. A useful agent is more than a webhook: durable execution, memory, file handling, approvals, tool access, logs, and failure recovery sit inside the same runtime.

Solution

Grail built a Rust-first runtime that receives messages, turns them into durable tasks, scopes the context and tools, runs the agent work, and posts results back to the original channel.

Impact

FastClaw became a base layer for Grail internal and client agents, supporting production-style work instead of simple chatbot replies.

Studio

MVPs for select entrepreneurs at lightning speed

We work with select entrepreneurs to build consumer products quickly using Grail's internally built Finite Machine system for app generation, review, preview, and launch.

Studio

Lobang property matching agent

A WhatsApp-first studio product that collects listings and buyer requirements, verifies agent details, filters unsafe messages, scores matches, and reveals contact links only after both sides confirm.

Read case study

Problem

Property agents already exchange buyer requirements and listings in WhatsApp, but manual matching is messy. A useful consumer workflow had to verify agents, filter irrelevant or unsafe messages, score matches, and protect trust before contact details were shared.

Solution

Grail built a headless WhatsApp agent that parses listings and requirements, verifies CEA details, checks relevance and safety, scores matches, asks both sides to confirm, and only then reveals contact details.

Stack used

Temporal

Postgres

WhatsApp Cloud APIEvolution/BaileysFlue

Slack admin

Impact

The prototype demonstrates a WhatsApp-native marketplace workflow with guardrails and double opt-in for contact sharing.

Studio

Astrology app

A consumer astrology MVP shaped through Grail Studio: concept, generated interface, lightweight content flow, and a usable web experience for early testing.

Read case study

Problem

Consumer products need fast taste-making before heavy engineering. The useful question was whether an astrology idea could become a polished first experience quickly enough to test positioning, flow, and user interest.

Solution

Grail used the studio app-generation workflow to turn the concept into a frontend experience, then reviewed the flow as a lightweight consumer MVP rather than a long product build.

Stack used

Finite MachineStudio buildFrontend appGenerated UI

Impact

The app became a small proof of how quickly the studio can move from consumer concept to usable web experience.

Studio

Vibe coding platform

A retail-focused app-building platform for turning product prompts into working app surfaces with sandboxed generation, preview links, deployable artifacts, and the internal Finite Machine system.

Read case study

Problem

Retail use cases need working product surfaces quickly, but normal app builds can take too long before the idea is concrete enough to test. Grail needed a faster way to generate, edit, preview, and package applications from prompts.

Solution

The team built a prompt-to-app workflow for retail use around sandboxed development environments, generated app surfaces, preview links, and deployable artifacts.

Impact

At launch, the platform reached 100 weekly active users and gave the team a real usage loop for prompt-to-app building.