[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"profile-en":3,"$f3e63a85gt5ru4":74},{"fullName":4,"role":5,"headline":6,"intro":7,"about":8,"essay":9,"location":10,"currentPosition":11,"availability":12,"email":13,"phone":14,"whatsappUrl":15,"linkedinUrl":16,"githubUsername":17,"cvUrl":18,"photoUrl":19,"domains":20,"experiences":37,"testimonials":63,"updatedAt":72,"translations":73},"Marc Ariel Oka","Senior Software Engineer · Fullstack","I design and build software products, *from the system to the experience.*","Senior fullstack software engineer based in Abidjan. I design and build complete web and mobile applications, with a frontend specialisation and projects in artificial intelligence, automation and machine learning. Founder of FACTio, an invoicing SaaS in production.","I am a fullstack software engineer. I design and build complete applications, web and mobile: architecture, business logic, APIs, databases and interfaces.\n\n## How I approach a problem\n\nBefore writing code, I try to understand the business: who uses the tool, what constraints they work under, what absolutely has to work. A good part of my work is turning scattered processes — spreadsheets, WhatsApp threads, manual data entry — into reliable, clear applications.\n\n## What I do\n\nI work on the backend as much as on the frontend, mainly with Laravel, Vue.js, Nuxt and Flutter. I am currently a senior fullstack developer at Flot, with solid frontend experience. Outside that role, I work on artificial intelligence, automation and machine learning projects in Python, sometimes with models running locally through Ollama.\n\n## FACTio, my product\n\nI designed, built and shipped FACTio, my own SaaS for invoicing and payment tracking, aimed at freelancers and small businesses. I keep evolving it based on what its users need.\n\n## What I want to go deeper into\n\nAI engineering and machine learning applied to real products, the architecture of applications that need to last, and the quality of interfaces on complex user journeys.","I work across the whole of an application, web or mobile: its architecture, business logic, APIs, data and interfaces. I am currently a senior fullstack developer at Flot.\n\nWhat I am most passionate about is applied research in artificial intelligence: understanding how models work, putting them to the test on real problems and seeing what holds up in production. I explore agents, automation and machine learning, often with models running locally.\n\nI am also building [FACTio](\u002Fprojets\u002Ffactio), my own invoicing SaaS, now in production. Along the way, I have designed platforms such as [Perafind](\u002Fprojets\u002Fperafind), [DS-Accessories](\u002Fprojets\u002Fds-accessories) and [Digifop](\u002Fprojets\u002Fdigifop) for AGEFOP.","Abidjan, Côte d’Ivoire","Senior Fullstack Developer at Flot","Available for projects, locally and internationally","okaamarc@gmail.com","+225 05 76 10 52 39","https:\u002F\u002Fwa.me\u002F2250576105239","https:\u002F\u002Fwww.linkedin.com\u002Fin\u002Fmarc-ariel-oka\u002F","Marc-Ariel-Oka","\u002Fcv\u002Fmarc-ariel-oka-cv.pdf","\u002Fimages\u002Fmarc-ariel-oka.webp",[21,25,29,33],{"title":22,"tools":23,"problem":24},"Fullstack Software Engineering","Laravel · PHP · PostgreSQL · MySQL · Docker","Designing a complete application — architecture, business logic, API, database, interface — and evolving it in production.",{"title":26,"tools":27,"problem":28},"Advanced Frontend","Vue.js · Nuxt · TypeScript · Tailwind CSS · Flutter","Fast, readable interfaces, including on complex journeys, on the web as well as on mobile.",{"title":30,"tools":31,"problem":32},"AI Engineering & Automation","Python · Ollama · Agents IA","Automating repetitive tasks and connecting language models to your data — locally when confidentiality requires it.",{"title":34,"tools":35,"problem":36},"Machine Learning","Python","Designing and integrating models when hand-written rules are no longer enough.",[38,44,49,54,59],{"title":39,"period":40,"company":41,"current":42,"description":43},"Senior Fullstack Developer","2025 — present","Flot",true,"Implementing designs and building high-performance interfaces, creating APIs and integrating third-party services. Leading integrations and the production releases of the web and mobile applications.",{"title":45,"period":46,"company":47,"description":48},"Fullstack Developer","2025","Emeryte","Building internal software: a money transfer application and a sports betting management platform.",{"title":50,"period":51,"company":52,"description":53},"Fullstack & IoT Developer","2024 — 2025","Orange Digital Center","Selected for the Orange Digital Academy programme to design an innovative project combining IoT and artificial intelligence.",{"title":55,"period":56,"company":57,"description":58},"Web Developer","2022 — 2024","In-novation","Building a platform to manage paid assignments for influencers and content creators.",{"title":60,"period":61,"company":57,"description":62},"Web Developer Intern","2022","Contributed to building an e-commerce template generator for small retailers.",[64,68],{"role":65,"quote":66,"author":67},"Project Manager — Perafind","Marc Ariel Oka est un travailleur vraiment sérieux. En travaillant avec lui sur Perafind, j'ai apprécié son dévouement et sa passion pour son métier de développeur. Dynamique et ouvert d'esprit, il sait piloter un projet et aime le travail bien fait. Je suis très satisfait de collaborer avec lui.","Franck Joris Anael Ahoussi",{"role":69,"quote":70,"author":71},"CEO — DS-Accessories","Un travail vraiment impeccable, rien à redire.","Michèle Samuela Koumi Ntakpe","2026-11-05T20:00:57.915801+00:00",{},{"post":75,"html":115,"toc":116,"previous":149,"next":152},{"id":76,"slug":77,"title":78,"excerpt":79,"content":80,"coverUrl":81,"coverAlt":81,"status":82,"featured":42,"readingMinutes":83,"lang":84,"translated":42,"seoTitle":85,"seoDescription":86,"translations":87,"publishedAt":88,"updatedAt":89,"category":90,"tags":95},"5e9eb460-ea20-4fd9-aac4-c9ae5c665c0b","de-l-automatisation-a-l-autonomie-ce-que-les-agents-ia-changent-pour-une-entrepr","From Automation to Autonomy: What AI Agents Change for a Business","Automating means running faster. Becoming autonomous means making better decisions. Here is the difference, and why it matters for an SME.","Picture a small business chasing unpaid invoices. It has automated the job: seven days after the due date, a reminder goes out on its own. That's already good. The time saved is real.\n\nBut the reminder also goes to the loyal client who always pays three days late. It goes to the client who announced a transfer yesterday. And nobody notices that late payments have been creeping up for two months.\n\nAutomation runs a rule. It never asks whether the rule still makes sense. That's where AI agents come in.\n\n## Automating is not the same as being autonomous\n\nAn automation follows a script written in advance: if this happens, do that. An AI agent receives a goal, looks at the situation and chooses the steps to reach it, using the tools it has been given access to.\n\n| | Automation | Autonomous agent |\n| --- | --- | --- |\n| Starting point | A rule | A goal |\n| Facing an unexpected case | Gets stuck or gets it wrong | Adapts, or asks a human |\n| Scope | One task, one tool | Several tools, sometimes several teams |\n| Afterwards | Does the same thing again | Takes into account what worked |\n| Stance | Reacts | Anticipates |\n\nIn an article on the subject, Nagarro's CTO, Rahul Mahajan, sums up what's at stake: automation brings efficiency, while autonomy makes it possible to decide based on context, collaborate across functions, learn from outcomes and anticipate.\n\n## What an agent adds\n\nBack to our invoice reminders.\n\n### 1. Deciding based on context\n\nThe agent checks the client's history before acting. The reliable payer who is a few days late gets a polite nudge. The client who has already announced a transfer gets nothing. The client who has ignored three reminders is flagged to a manager.\n\n### 2. Collaborating across functions\n\nAn unpaid invoice is not just an accounting matter. The agent can brief the sales rep before their next visit, or tell the logistics team that a delivery is waiting on a payment. Information flows without a meeting.\n\n### 3. Learning from outcomes\n\nThe agent keeps track of what works. If clients pay faster after a WhatsApp message than after an email, it notes it and adjusts. The automation, meanwhile, will send the same email for ten years.\n\n### 4. Anticipating\n\nBy following trends, the agent sees problems coming. Payment delays are growing for one type of client? It flags it before cash flow feels it. This is the shift from a company that reacts to one that anticipates.\n\n## Where companies really stand\n\nThings are moving fast. In August 2025, Gartner predicted that 40% of enterprise applications would feature task-specific agents by the end of 2026, up from less than 5% in 2025.\n\nBut Gartner also tempers the enthusiasm. It expects over 40% of agentic AI projects to be cancelled by the end of 2027, due to escalating costs, unclear business value or inadequate risk controls. It also warns about \"agent washing\": chatbots or classic automations rebranded as \"agents\". Out of thousands of vendors, it counts only about 130 offering genuine agents.\n\nIn other words, autonomy is not a product you buy. It's a capability you build, step by step.\n\n## For a small business: where to start\n\nYou don't need a multinational's budget. Above all, you need a method.\n\n1. **Pick one process that hurts.** Invoice reminders, WhatsApp orders, stock tracking, customer service. A repetitive process, with data, and a visible cost when it goes wrong.\n2. **Set a measurable goal.** \"Cut the average payment delay by 10 days\", not \"add some AI\".\n3. **Raise autonomy in stages.** First the agent suggests and a human approves. Then it acts alone on simple cases. It never decides alone on anything irreversible: a refund, a deletion, a financial commitment.\n4. **Log everything.** Every action the agent takes must be recorded: what it did, why, and with what data. Without a log, you can't fix anything.\n5. **Plan a stop button.** You should be able to switch the agent off in a minute without stopping the business.\n6. **Measure, then expand.** If the first agent delivers for a month or two, give it a second scope, or have it work with another team.\n\nOne last piece of advice: be wary of offers promising a turnkey \"agent\" without asking a single question about your processes. It's often an automation with a new name.\n\n## In short\n\nAutomation saves time. Autonomy adds judgement: deciding based on context, getting information to the right people, learning and seeing what's coming.\n\nAI agents don't replace your team. They take over the tasks that need no thinking, and bring the right information at the right time for the ones that do.\n\nIf you're wondering which of your processes could move from automation to autonomy, [get in touch](https:\u002F\u002Fwww.marc-oka.com\u002Fcontact).\n\n---\n\n### Sources\n\n- [From automation to autonomy: AI agents as the enterprise nervous system](https:\u002F\u002Fwww.nagarro.com\u002Ffr\u002Fblog\u002Fai-agents-for-enterprises-automation-to-autonomy), Rahul Mahajan, Nagarro, September 24, 2025 (French version).\n- [Gartner Predicts 40% of Enterprise Apps Will Feature Task-Specific AI Agents by 2026](https:\u002F\u002Fwww.gartner.com\u002Fen\u002Fnewsroom\u002Fpress-releases\u002F2025-08-26-gartner-predicts-40-percent-of-enterprise-apps-will-feature-task-specific-ai-agents-by-2026-up-from-less-than-5-percent-in-2025), Gartner, August 26, 2025.\n- [Gartner: More than 40% of agentic AI projects will be canceled in the next few years](https:\u002F\u002Fsdtimes.com\u002Fagentic-ai\u002Fgartner-more-than-40-of-agentic-ai-projects-will-be-canceled-in-next-few-years\u002F), SD Times, June 30, 2025.","","published",4,"en","AI Agents: From Automation to Autonomy for Small Businesses","Automation saves time. An autonomous AI agent decides, collaborates and learns. What it changes for a small business, and where to start.",{},"2026-11-14T08:00:00+00:00","2026-11-13T11:00:38.448113+00:00",{"id":91,"name":92,"slug":93,"description":94},"80f57329-8567-4a27-b0d9-138173db876b","Automation","automatisation","Automating business processes.",[96,100,104,107,111],{"id":97,"name":98,"slug":99},"2ad0af52-4df0-4b03-ac94-1ad2cca1617d","agents IA","agents-ia",{"id":101,"name":102,"slug":103},"0a7b766e-66de-425c-9f1e-ab93723f7d16","automatisation PME","automatisation-pme",{"id":105,"name":106,"slug":106},"f5d43502-8eae-487f-8eef-e7bf048884db","autonomie",{"id":108,"name":109,"slug":110},"0279027d-7eaa-436f-98ab-1c00fe171d77","Côte d'Ivoire","cote-d-ivoire",{"id":112,"name":113,"slug":114},"e0d508b2-5d4a-48fc-8bbb-311d8f8c6a9f","IA agentique","ia-agentique","\u003Cp>Picture a small business chasing unpaid invoices. It has automated the job: seven days after the due date, a reminder goes out on its own. That’s already good. The time saved is real.\u003C\u002Fp>\n\u003Cp>But the reminder also goes to the loyal client who always pays three days late. It goes to the client who announced a transfer yesterday. And nobody notices that late payments have been creeping up for two months.\u003C\u002Fp>\n\u003Cp>Automation runs a rule. It never asks whether the rule still makes sense. That’s where AI agents come in.\u003C\u002Fp>\n\u003Ch2 id=\"automating-is-not-the-same-as-being-autonomous\">Automating is not the same as being autonomous\u003C\u002Fh2>\n\u003Cp>An automation follows a script written in advance: if this happens, do that. An AI agent receives a goal, looks at the situation and chooses the steps to reach it, using the tools it has been given access to.\u003C\u002Fp>\n\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>\u003C\u002Fth>\n\u003Cth>Automation\u003C\u002Fth>\n\u003Cth>Autonomous agent\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003C\u002Fthead>\n\u003Ctbody>\n\u003Ctr>\n\u003Ctd>Starting point\u003C\u002Ftd>\n\u003Ctd>A rule\u003C\u002Ftd>\n\u003Ctd>A goal\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>Facing an unexpected case\u003C\u002Ftd>\n\u003Ctd>Gets stuck or gets it wrong\u003C\u002Ftd>\n\u003Ctd>Adapts, or asks a human\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>Scope\u003C\u002Ftd>\n\u003Ctd>One task, one tool\u003C\u002Ftd>\n\u003Ctd>Several tools, sometimes several teams\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>Afterwards\u003C\u002Ftd>\n\u003Ctd>Does the same thing again\u003C\u002Ftd>\n\u003Ctd>Takes into account what worked\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>Stance\u003C\u002Ftd>\n\u003Ctd>Reacts\u003C\u002Ftd>\n\u003Ctd>Anticipates\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftbody>\n\u003C\u002Ftable>\n\u003Cp>In an article on the subject, Nagarro’s CTO, Rahul Mahajan, sums up what’s at stake: automation brings efficiency, while autonomy makes it possible to decide based on context, collaborate across functions, learn from outcomes and anticipate.\u003C\u002Fp>\n\u003Ch2 id=\"what-an-agent-adds\">What an agent adds\u003C\u002Fh2>\n\u003Cp>Back to our invoice reminders.\u003C\u002Fp>\n\u003Ch3 id=\"1-deciding-based-on-context\">1. Deciding based on context\u003C\u002Fh3>\n\u003Cp>The agent checks the client’s history before acting. The reliable payer who is a few days late gets a polite nudge. The client who has already announced a transfer gets nothing. The client who has ignored three reminders is flagged to a manager.\u003C\u002Fp>\n\u003Ch3 id=\"2-collaborating-across-functions\">2. Collaborating across functions\u003C\u002Fh3>\n\u003Cp>An unpaid invoice is not just an accounting matter. The agent can brief the sales rep before their next visit, or tell the logistics team that a delivery is waiting on a payment. Information flows without a meeting.\u003C\u002Fp>\n\u003Ch3 id=\"3-learning-from-outcomes\">3. Learning from outcomes\u003C\u002Fh3>\n\u003Cp>The agent keeps track of what works. If clients pay faster after a WhatsApp message than after an email, it notes it and adjusts. The automation, meanwhile, will send the same email for ten years.\u003C\u002Fp>\n\u003Ch3 id=\"4-anticipating\">4. Anticipating\u003C\u002Fh3>\n\u003Cp>By following trends, the agent sees problems coming. Payment delays are growing for one type of client? It flags it before cash flow feels it. This is the shift from a company that reacts to one that anticipates.\u003C\u002Fp>\n\u003Ch2 id=\"where-companies-really-stand\">Where companies really stand\u003C\u002Fh2>\n\u003Cp>Things are moving fast. In August 2025, Gartner predicted that 40% of enterprise applications would feature task-specific agents by the end of 2026, up from less than 5% in 2025.\u003C\u002Fp>\n\u003Cp>But Gartner also tempers the enthusiasm. It expects over 40% of agentic AI projects to be cancelled by the end of 2027, due to escalating costs, unclear business value or inadequate risk controls. It also warns about « agent washing »: chatbots or classic automations rebranded as « agents ». Out of thousands of vendors, it counts only about 130 offering genuine agents.\u003C\u002Fp>\n\u003Cp>In other words, autonomy is not a product you buy. It’s a capability you build, step by step.\u003C\u002Fp>\n\u003Ch2 id=\"for-a-small-business-where-to-start\">For a small business: where to start\u003C\u002Fh2>\n\u003Cp>You don’t need a multinational’s budget. Above all, you need a method.\u003C\u002Fp>\n\u003Col>\n\u003Cli>\u003Cstrong>Pick one process that hurts.\u003C\u002Fstrong> Invoice reminders, WhatsApp orders, stock tracking, customer service. A repetitive process, with data, and a visible cost when it goes wrong.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Set a measurable goal.\u003C\u002Fstrong> « Cut the average payment delay by 10 days », not « add some AI ».\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Raise autonomy in stages.\u003C\u002Fstrong> First the agent suggests and a human approves. Then it acts alone on simple cases. It never decides alone on anything irreversible: a refund, a deletion, a financial commitment.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Log everything.\u003C\u002Fstrong> Every action the agent takes must be recorded: what it did, why, and with what data. Without a log, you can’t fix anything.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Plan a stop button.\u003C\u002Fstrong> You should be able to switch the agent off in a minute without stopping the business.\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Measure, then expand.\u003C\u002Fstrong> If the first agent delivers for a month or two, give it a second scope, or have it work with another team.\u003C\u002Fli>\n\u003C\u002Fol>\n\u003Cp>One last piece of advice: be wary of offers promising a turnkey « agent » without asking a single question about your processes. It’s often an automation with a new name.\u003C\u002Fp>\n\u003Ch2 id=\"in-short\">In short\u003C\u002Fh2>\n\u003Cp>Automation saves time. Autonomy adds judgement: deciding based on context, getting information to the right people, learning and seeing what’s coming.\u003C\u002Fp>\n\u003Cp>AI agents don’t replace your team. They take over the tasks that need no thinking, and bring the right information at the right time for the ones that do.\u003C\u002Fp>\n\u003Cp>If you’re wondering which of your processes could move from automation to autonomy, \u003Ca href=\"https:\u002F\u002Fwww.marc-oka.com\u002Fcontact\" target=\"_blank\" rel=\"noopener noreferrer\">get in touch\u003C\u002Fa>.\u003C\u002Fp>\n\u003Chr>\n\u003Ch3 id=\"sources\">Sources\u003C\u002Fh3>\n\u003Cul>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fwww.nagarro.com\u002Ffr\u002Fblog\u002Fai-agents-for-enterprises-automation-to-autonomy\" target=\"_blank\" rel=\"noopener noreferrer\">From automation to autonomy: AI agents as the enterprise nervous system\u003C\u002Fa>, Rahul Mahajan, Nagarro, September 24, 2025 (French version).\u003C\u002Fli>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fwww.gartner.com\u002Fen\u002Fnewsroom\u002Fpress-releases\u002F2025-08-26-gartner-predicts-40-percent-of-enterprise-apps-will-feature-task-specific-ai-agents-by-2026-up-from-less-than-5-percent-in-2025\" target=\"_blank\" rel=\"noopener noreferrer\">Gartner Predicts 40% of Enterprise Apps Will Feature Task-Specific AI Agents by 2026\u003C\u002Fa>, Gartner, August 26, 2025.\u003C\u002Fli>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fsdtimes.com\u002Fagentic-ai\u002Fgartner-more-than-40-of-agentic-ai-projects-will-be-canceled-in-next-few-years\u002F\" target=\"_blank\" rel=\"noopener noreferrer\">Gartner: More than 40% of agentic AI projects will be canceled in the next few years\u003C\u002Fa>, SD Times, June 30, 2025.\u003C\u002Fli>\n\u003C\u002Ful>\n",[117,121,124,128,131,134,137,140,143,146],{"id":118,"text":119,"level":120},"automating-is-not-the-same-as-being-autonomous","Automating is not the same as being autonomous",2,{"id":122,"text":123,"level":120},"what-an-agent-adds","What an agent adds",{"id":125,"text":126,"level":127},"1-deciding-based-on-context","1. Deciding based on context",3,{"id":129,"text":130,"level":127},"2-collaborating-across-functions","2. Collaborating across functions",{"id":132,"text":133,"level":127},"3-learning-from-outcomes","3. Learning from outcomes",{"id":135,"text":136,"level":127},"4-anticipating","4. Anticipating",{"id":138,"text":139,"level":120},"where-companies-really-stand","Where companies really stand",{"id":141,"text":142,"level":120},"for-a-small-business-where-to-start","For a small business: where to start",{"id":144,"text":145,"level":120},"in-short","In short",{"id":147,"text":148,"level":127},"sources","Sources",{"slug":150,"title":151},"les-failles-explosent-ce-que-doit-vraiment-faire-une-pme","Les failles explosent: ce que doit vraiment faire une PME",{"slug":153,"title":154},"monolithe-ou-microservices-ce-qu-une-pme-devrait-vraiment-choisir","Monolith or Microservices: What a Small Business Should Really Choose"]