World Health Organization

Compliance and Risk Management Officer

Quality Assurance Specialist- African Regional Office

UNDP

Operations Officer

MONUSCO

Budget Specialist

Finance Associate

Merck

Finance Manager

Schering-Plough Research Institute

Assistant Manager Finance

Auditor

Empowering non-Tech Founders, Solopreneurs, Professionals, online businesses with ChatGPT & AI to supercharge productivity | ex. EY | serving UN | CPA, FCCA | Follow to learn 👇

I am an experienced Compliance and Risk Management Professional with over a decade of expertise in international operations, financial management, and quality assurance. Currently serving as a Compliance and Risk Management Officer at the World Health Organization, I bring a proven track record of fostering operational excellence, implementing robust risk frameworks, and ensuring organizational compliance across diverse regions.

What Sets Me Apart:

- Holistic Approach to Leadership: Beyond professional success, I believe in integrating personal transformation and emotional resilience into my work. 

- Mentorship & Growth: I am passionate about coaching and guiding professionals, helping them unlock their potential in both established and emerging fields like AI and personal development.

- Core Values: Transparency, openness, and trust form the foundation of my professional and personal journey.

Personal Philosophy:

Balancing the technical with the transformative, I strive to bridge the gap between operational efficiency and personal growth. Whether leading risk mitigation strategies or exploring the art of feminine embodiment, I am committed to creating a legacy that inspires alignment, authenticity, and abundance in every sphere of life.

Outside of my professional work, I am a lifelong learner, an AI enthusiast, and a proponent of holistic growth—committed to harmonizing professional excellence with personal fulfillment.

Let’s connect to explore how we can create impactful change together!

Beenish Arif

DreamGrow Digital

I am responsible for SEO, content marketing, Google Ads and Analytics. I manage strategic planning, key client and partner relations, marketing, training, and consulting.

Founder, Strategist

Äripäev

A radio show about practical sides of internet marketing.

Radio Talk Show Host

Enterprise Estonia

Business Mentor

Tallinn Business Incubators

Mentoring incubants in Loomeinkubaator. Creating seminars and workshops to help entrepreneurs get their marketing on track.

Freelancer

I have written thousands of articles, blog posts and e-books. The largest portion of that material covers digital marketing and management. I have also covered social issues, motivation, psychology, and science. I write in English and Estonian.

Best results include social media marketing blog with more than 140k visitors per month and millions of total readers.

Blogger, writer, journalist

Sevenline

Internet marketing and strategy for key clients.

Shipitwise

Now you can send your bulky purchases home using a simple app on your phone. We will deliver large items that are too inconvenient for air travel directly to your home address. You can even use 2-way shipping for your sporting gear and outdoor equipment.

Chief Marketing Officer

NetLead OÜ

Launching and staffing an international SEO company. Activities in 10 countries.

Altex Marketing

Managing day to day business. Internet marketing strategies and implementation for the clients. Launching a series of internet marketing seminars in Estonia.

Member of the Board

Infinit Solutions

Managing the daily activities of the company. Internet marketing strategy and consultation for the clients.

Ole internetis leitav ja too kliendid oma keskkonda. Optimeerimine, koolitused, konsultatsioonid. Küsi, kuidas tulemusi saada

I help your team to improve productivity with AI tools.

AI transformation for businesses & entrepreneurs

✔️ Organized the first series of AI seminars in Estonia
✔️ Helped more than 10,000 professionals with AI productivity
✔️ Spoken at 700+ seminars and conferences on AI, strategy, sales, and marketing
✔️ Lecturer in Estonian universities
✔️ Mentoring business leaders and entrepreneurs marketing, productivity, and AI

Day to day

✔️ Using AI to improve productivity
✔️ Training and consulting
✔️ Key client and partner relations
✔️ SEO and content strategy

Helping clients succeed

✔️ 27+ years managing marketing projects and business development
✔️ Providing marketing services since 1995, helping companies reach clients and increase sales
✔️ Founder of the first digital marketing agency in Estonia
✔️ Project management experience in sales, marketing, and advertising
✔️ Participated in a wide array of software and internet projects, from content management systems to server software

Publications

✔️ Numerous articles in the largest Estonian newspapers
✔️ Countless articles (1500+) about marketing, sales, productivity, and AI
✔️ Built digital marketing blog Dreamgrow.com to more than 250,000 visitors/month

Specialties: AI tools and productivity, business systems, digital strategy, inbound marketing, analytics, SEO, content marketing

Priit Kallas

Google #NanoBanana Pro on väga raju asi.

Varasematel mudelitel on olnud raskusi teksti korralikult kirjutamisega hakkama saada.

Nüüd Google DeepMind viimane pildimasin suudab teha korraliku teksti, mõelda välja loogilise ahela, seda pildil kujutada... ja samal ajal ka päriselt nalja teha (vaataja nalja tunnetamise tase ei pruugi olla lisatud pildiga kompatiibel)

Alguses ei saanud pildimasinad nagu Midjourney isegi lihtsa teksti kirjutamisega hakkama. Siis tuli OpenAI pildi #AI, see sai enamus tekstiga enamvähem hakkama. Samas eesti keeles Õ oli talle pea ületamatuks probleemiks.

Vähem kui aastaga on pildi AI jõudnud tasemele, kus teksti saab pildile peaaegu veatult. Peamine põhjus on uute mudelite uus tehnoloogia.

Midjourney (difusioon) on nagu impressionistlik kunstnik. Annad teema ja masin hakkab lõuendile värvi loopima. Silub seda seni, kuni see näeb ilus välja. Tulemus on emotsionaalne ja stiilne, aga kui sa palud tal maalida täpse vöötkoodi, teksti või logo, siis tulemus ei ole väga hea, sest ta maalib "tunde järgi".

Nano Banana (uus transformer mudel) on nagu insener või printer. Masinal on hiiglaslik kogu valmis elementidest ja ta saab aru struktuurist. Kui palud teha plakati, paigutab ta elemendid loogiliselt paika. Ta "saab aru", mis on "särk" ja mis on "tekst särgil", ega sulata neid kogemata kokku. See annab ka kiirema tulemuse ja kulutab vähem energiat. Miinus: Tulemus võib vahel tunduda natuke "plastmass" või liiga puhas, sest see on kokku pandud loogilistest plokkidest, mitte orgaanilisest mürast.

Midjourneyt ei peaks siiski maha kandma. see on minu arust siiani kõige parema "kunstimaitsega" pildi loomise AI.

Kus sa kasutad oma protsessides AI-piltide tegemist?

PS Jälgi mind, kui tahad teada, milliseid asju pead tegelikult AI teemadest teadma.

Priit Kallas ✨

AI productivity boost for your team. AI automation, consulting, training. Find out how to 3x your productivity.

Gemini 3 tuli eile välja. Kui GPT-5.1 ja #Claude 4.5 tundusid paigalseisuna, siis nüüd on tunda teatavat edasiminekut.

#Gemini 3 LMArena Elo skoor 1501 on suur samm edasi võrreldes just avaldtatud GPT ja Groki mudelitega.

Teksti kvaliteedi osas on hea asi uus "Thinking Level" parameeter. Kui keerad selle põhja, kaob see tüütu tehisintellekti pugemine ja ülemäärane viisakus ära. Selle asemel saad tipptasemel teksti. Gemini 3 kirjutab nii, et tulemuseks on sügav arusaam teemast, mitte mull

"Vibe Coding": Uus Gemini 3 peal toimiv Antigravity platvorm ei kirjuta lihtsalt koodijuppe, vaid käitub arhitektina ning kasutab haldab terminali, redaktorit ja brauserit korraga.

Humanity's Last Exam (HLE) tulemus 37.5%. Kontekstiks: Gemini 2.5 sai seal 21.6% ja Claude Sonnet 4.5 vaid 13.7%.

Nagu varasemalt on Gemini 3 miljoni tokeni suurune kontekstiaken. Nii saad normaalseid tulemusi ka suuremate andmemahtudega. Treeningandmed ulatuvad 2025. aasta jaanuari.

Hind ei ole odav, aga selle võimekuse eest väga hulle ei ole:
Sisend: $2.00 / 1 miljon tokenit.
Väljund: $12.00 / 1 miljon tokenit (kuni 200k pikkuste promptide puhul).

Samas koodi võimekuse osas on mõned küsitavused, SWE-bench edetabelis on mudel alles 3. kohal, mis on spetsiifiliste bugide parandamisel ohumärk. Ka matemaatikaga on segased lood, mõned saavad väha häid tulemusi, teised ainult hallukaid.

Ma ise kasuta üha rohkem oma töös Geminit, sest eesti keel tundub soravam ja vastused põhjalikumad. Eriti kasulik kohtades, kus vaja pikka sisulist vastust ja on väga palju andmeid sisendiks. Google DeepMind #AI on ka otsingu kasutamisel ootuspäraselt väga hea.

Mis mudelit sa ise igapäevaselt oma töös kasutad?

PS Jälgi mind, kui tahad teada, milliseid asju pead tegelikult AI teemadest teadma.

Yesterday I experimented with the new #Gemini 3 Pro model. I used it for writing, coding, and experimented with images.

The results are great!

First, the most trivial thing. Writing with Google DeepMind's Gemini 3 is great! However, if you have used 2.5 for some time then there's not a huge leap. It is great but you already got good enough content for #sales, #marketing, and #writing from Gemini 2.5 pro.

Next, images with #NanoBanana Pro. This is next level! There will be so much work you can do with images now that were IMPOSSIBLE 2 weeks ago.

Finally, I moved on to #vibecoding. I needed to finish one of my customer's project. The project involves everything you would expect from a business #automation workflow:

1. Getting the information,

2. enriching the collected data (one part of it is using #AI vision to get data from photos),

3. adding more information from public sources, and finally

4. delivering the result in an appealing web UI.

I worked on this project for about 4 hours during the day. But the impressive part is that AI worked for me

1 hour, 44 minutes, and 29 seconds!!!

Gemini 3 Pro wrote 3061 lines of code. If you consider 100 lines per hour as an average dev speed, this means 30 HOURS!!! of dev work. 4 days! Non-stop!

But that's not all, I also wrote about 1,500 lines of HTML. This is usually done with HTML-editors so let's say 2 to 3 additional hours.

So, in total AI did the work of 4 full days in less than 2 hours. If I had used a human for this the code would be ready some time in the next 10 days!!!

No, it was not the most complicated code, but is solves a problem. Before AI this work would have taken at minimum 2 weeks and the invoice from the dev would have been at THE VERY LEAST in the 3000+ range.

Now it took me a DAY!!!

And remember, this is the worst AI will ever be at coding. New models come out every couple of months and the improvements are huge!

What's your biggest win with AI?

PS Follow me for more info about using #LLMs in business either for productivity or defending from attacks.

#NanoBanana PRO is a powerful! It has both image editing and creation

Here are some practical ways you can use it to in your workflow:

1. Maintain visual consistency for branding and content
Character and model consistency: Use a reference image of a person or character and place them into entirely new scenes (a coffee shop, a detailed setting) and keeping their look identical. excellent for consistent advertising campaigns.

Product placement: add products into existing images. For example, replace a generic object in a person's hand with a specific branded item, like swapping a drinks can for a specific bottle.

2. Rapid scene and environment modification
Interior design: Transform interior spaces. Take a photo of a room and prompt it to become a refurbished, modern room keeping the original dimensions intact. Then, add furniture, plants, or change the lighting (setting the scene to night).

Weather and lighting adjustments: Change the atmosphere of a photo by prompting specific weather conditions (heavy rain, thick snow, sandstorm) or adjusting lighting for day-to-night effects.

Selective style changes: Apply different art styles (like N64 graphics or impressionist painting) to an entire image, or apply a style to only a specific object within the photo (making a piece of food look like a hand-drawn illustration while keeping the rest realistic).

3. Efficient editing and composition control
Clean up backgrounds: remove unwanted objects from a scene, such as lines or distracting elements and keep the subject the same.

Control camera angles: Generate different perspectives from a single input image. Transforming a ground-level street shot into an aerial or bird’s-eye view. This helps simulate multiple camera setups without needing new photos.

4. Restoration and historical updates
Restore damaged or faded vintage family photos using a simple text prompt.

Modernize imagery: add color to old photos, apply natural skin tones, and update a scene by placing the historical subject into a contemporary settin.

Do you have any tasks in your workflow where you can use Google DeepMind's new image #AI?

PS Follow me for more info about using #LLMs in business either for productivity or defending from attacks.

Eile tulid NVIDIA järjekordsed tulemused. Järekordselt analüütikud alahindasid nõudlust #AI riistvara vastu.

AI võib olla hetkeline mull, #NVDA võib ka põhja minna, aga sarnaselt arvuti ja interneti tulemisega muutub see meie elu osaks ja mõjutab sadu triljoneid globaalset GDPd järgmiste aastate jooksul

See on põhjus, miks suured tegijad praegu näiliselt raha põletavad.

Mäng käib 35 triljoni dollari peale aastas. Peaga töötajate osa maailma GDPs on ca kolmandik. Maailma GDP on ligikaudu 110 triljonit dollarit.

See on summa, mida maailm maksab igal aastal inimestele "peaga tehtava" töö eest. See number paneb kõik hiljutised Nvidia tulemused ja OpenAI investeeringud hoopis teise perspektiivi.

Selline matemaatika:

Turg on hoomamatult suur. 35 triljonit dollarit aastas läheb praegu inseneridele, analüütikutele, kirjutajatele ja juhtidele. Tehnoloogiafirmad ei sihi enam ainult reklaamiturgu, vaid globaalset palgafondi.

Investeeringud on selle kõrval peenraha. Kui kuuled, et keegi ehitab 100 miljardi eest andmekeskust, siis see tundub kallis. Aga kui see annab ligipääsu turule, mille maht on 35 000 miljardit, on see soodusmüük.

Eesmärk on kognitiivne tööjõud. Jutt ei käi enam paremast otsingumootorist. Jutt käib tarkvarast, mis suudab teha ära keskmise kontoritöötaja töö.

Marginaalid muutuvad. Inimtööjõu skaleerimine on kallis ja aeglane. Digitaalse "mõttetöö" paljundamine maksab murdosa sellest. See on põhjus, miks investorid ei vaata praeguseid kulusid, vaid tuleviku potentsiaali.

Siin on üks kainestav reaalsuskontroll. See üleminek ei juhtu üleöö ega nipsust. Füüsiline maailm seab piirid, meil pole täna veel piisavalt ressursse, et jooksutada mudeleid, mis suudaksid kogu seda 35 triljonilist koormust kanda.

Aga suund on selge.

Kas sinu valdkonnas on juba tunda, et AI hakkab päriselt "mõttetööd" üle võtma või on see endiselt pigem lihtsalt abimees?

PS Jälgi mind, kui tahad teada, milliseid asju pead tegelikult AI teemadest teadma.

taplio

The Battle

Beenish Arif

Priit Kallas

Followers

Posts

Engagements

Likes

Performance ScorePriit Kallas's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Beenish Arif

Priit Kallas

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Priit Kallas's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScorePriit Kallas's LinkedIn Performance