talalkhawaja.com · teqprotech.com · LinkedIn · Devpost · lablab.ai · ORCID
I'm Talal Khawaja, a Production Manager and digital systems builder. I build systems that keep working without supervision: Laravel platforms, CRM and AI automation, API integrations and the technical operations that keep them running. Where a decision matters, the AI proposes and a person approves.
- Production Manager (CRM & automation operations) since March 2025: GoHighLevel automation, AI booking bots, integrations, email deliverability and web builds.
- Founder of Teqprotech since September 2021, a small remote web-development studio.
- Full-stack developer (PHP / Laravel) since 2020.
| Area | In practice |
|---|---|
| Laravel & PHP | Production web platforms. talalkhawaja.com and teqprotech.com run on one domain-aware Laravel 12 app with a Filament admin. |
| Python | FastAPI services and agents (ThesisCircuit, NeighborOps AI), plus automation scripts. |
| GoHighLevel & CRM | Workflows, pipelines, AI booking bots, lead routing and reporting for day-to-day operations. |
| APIs & integrations | Connecting CRMs, forms, telephony, email and AI services so data moves without copy-paste. |
| AI automation | Agents that call deterministic tools and stop for a person at the decisions that matter. |
| Workflow automation | Triggers, schedules and hand-offs that remove repetitive work from a team's day. |
| WordPress | Business sites, maintenance and fixes. |
| Application development | TypeScript and Next.js builds, from hackathon prototypes to deployed demos. |
| Technical operations | CI/CD with GitHub Actions, backups and health-checked deploys, email deliverability, monitoring. |
Finalist — USAII Global AI Hackathon 2026
An AI decision-support dashboard for community AI readiness. Transparent local scoring finds the gaps, scenarios can be compared side by side, and the output is a three-phase responsible-AI roadmap with explicit human decision points. The model advises; people decide.
Stack: Next.js, TypeScript, OpenAI SDK, Recharts · Last updated: 28 Sep 2026
Repository · Case study · Live demo · Devpost
- Finalist — USAII Global AI Hackathon 2026 for CivicAI Readiness Blueprint: AI decision-support dashboard for community AI readiness, with a three-phase responsible-AI roadmap and human decision points.
| Platform | Snapshot | Last verified |
|---|---|---|
| lablab.ai | Elite level · 3,865 points · Leaderboard #15 · 10 events · 5 submissions | 2026-09-30 |
| Devpost | 6 hackathons · 8 projects | 2026-09-27 |
Platform numbers are dated snapshots, not live counters.
Hackathon entries (2026)
| Project | Event | Result | Team | Links |
|---|---|---|---|---|
| KhidmatConnect AI | Alibaba Cloud AI Hackathon Pakistan 2026 (Open Innovation) · Organiser portal · 3 Oct 2026 | Participation — Karachi Regional Technical Evaluation | Team lead, with Aqeela Urooj | Repo · Case study · Certificate |
| LockSmith | IBM Bob 2.0 Hackathon · lablab.ai · 27 Sep 2026 | Submitted, results pending | Solo | Repo · Case study · Submission |
| DUET | AI Infra Summit Hackathon · lablab.ai · 16 Sep 2026 | Submitted | Team of 4 | Repo · Case study |
| NeighborOps AI | Agents for Humans Hackathon (Good Neighbor Agents track) · Devpost · 14 Sep 2026 | Submitted, results pending | Team of 4, with Aqeela Urooj | Repo · Case study · Submission |
| ThesisCircuit | Alpaca AI Trading Agents Hackathon · lablab.ai · 4 Sep 2026 | Submitted | Team | Repo · Case study |
| ClientOps Memory AI | CockroachDB × AWS Hackathon: Build with Agentic Memory · Devpost · 15 Aug 2026 | Submitted | With Aqeela Urooj | Case study · Submission |
| LaunchMap AI | AI Factory (native.builder) Hackathon · lablab.ai · 7 Aug 2026 | Submitted | With Aqeela Urooj | Repo · Case study · Submission |
| BrandPilot Local | AMD AI DevMaster Hackathon (Track 1, Multimodal) · GitHub · 5 Aug 2026 | Submitted | With Aqeela Urooj | Repo · Case study · Submission |
| QuotePilot AI | Global AI Hackathon Series with Qwen Cloud (Track 4) · Devpost · 19 Jul 2026 | Submitted | With Aqeela Urooj | Repo · Case study · Submission |
| ScopeGuard AI | OpenAI Build Week (Work & Productivity) · Devpost · 18 Jul 2026 | Submitted | With Aqeela Urooj | Repo · Case study · Submission |
| CaptionForge AI | AMD Developer Hackathon: ACT II (Track 2) · lablab.ai · 13 Jul 2026 | Submitted | Team of 6, with Aqeela Urooj | Repo · Case study |
| CivicAI Readiness Blueprint | USAII Global AI Hackathon 2026 · Devpost · 21 Jun 2026 | Finalist | With Aqeela Urooj | Repo · Case study · Submission |
Certificates
- Certificate of Participation — Karachi Regional Technical Evaluation · Alibaba Cloud AI Hackathon Pakistan 2026 · KhidmatConnect AI (P00542, Open Innovation), NASTP Karachi · 3 Oct 2026 (last verified 2026-10-08)
- Certificate of Completion (Participant) · lablab.ai · IBM Bob 2.0 hackathon · 28 Sep 2026 (last verified 2026-09-30)
- Certificate of Completion (Participant) · lablab.ai · AI Infra Summit Hackathon · 22 Sep 2026 (last verified 2026-09-30)
- Certificate of Completion (Participant) · lablab.ai · Alpaca AI Trading Agents Hackathon · 14 Sep 2026 (last verified 2026-09-30)
- Certificate of Completion (Participant) · lablab.ai · AMD Developer Hackathon: ACT II · 13 Aug 2026 (last verified 2026-09-30)
- Certificate of Completion (Participant) · lablab.ai · AI Factory · 13 Aug 2026 (last verified 2026-09-30)
- Let the model reason, not promise: building QuotePilot AI · 7 Oct 2026
How QuotePilot AI keeps Qwen away from prices: schema-checked analysis, deterministic quoting tools and a mandatory human approval step before anything is sent. - What LockSmith checks before a PostgreSQL migration ships · 7 Oct 2026
The PostgreSQL migration risks LockSmith flags, how it proves each lock in embedded Postgres, and why its blocking-time numbers are labelled as estimates.
| Repository | What it is | Language | Updated |
|---|---|---|---|
| dark-factory-pocketful | A four-seat BAND Desktop dark factory: one task message → pocketful payments service, stages 1–4, built, reviewed and verified autonomously. | JavaScript | 5 Oct 2026 |
| scopeguard-ai · demo | Evidence-backed scope analysis and human approval workspace, built with Codex using GPT-5.6. | TypeScript | 28 Sep 2026 |
| neighborops-ai · demo | Community-pantry coordination: a Strands agent handles routine requests while staff decide how scarce resources are used. Agents for… | Python | 28 Sep 2026 |
| quotepilot-ai · case study | Autonomous quotation and customer-intake agent powered by Qwen Cloud and Alibaba Cloud Function Compute | TypeScript | 28 Sep 2026 |
| thesiscircuit · demo | Paper-only autonomous options agents with deterministic risk vetoes and replayable evidence. | Python | 28 Sep 2026 |
| locksmith-ibm-bob · demo | LockSmith — catch PostgreSQL migrations that lock production; IBM Bob custom mode + parallel subagents rewrite them into verified… | TypeScript | 28 Sep 2026 |
More write-ups and case studies: talalkhawaja.com/projects
- Deterministic first, model second. Prices, limits and risk checks live in code; the model reasons around them.
- A person approves what matters. Quotes, scarce resources and client delivery plans wait for a human decision.
- Evidence over claims. Every project links to its code, a working demo where one exists, and a written case study.
Business enquiries: info@teqprotech.com · Portfolio: talalkhawaja.com
The Featured, Achievements, Latest writing and Recently updated sections are refreshed daily by a GitHub Action from verified data, the talalkhawaja.com sitemap and the GitHub API.